Pith. sign in

REVIEW 3 major objections 5 minor 255 references

Slot2Text: Object-Centric Visual Tokenization for Efficient and Spatially Traceable Surgical MLLMs

T0 review · 3 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Slot2Text replaces 1,295 dense visual tokens with 47 slot tokens in a surgical MLLM, staying competitive.

desk verdict A genuinely useful token-compression interface for surgical MLLMs, with a credible efficiency story and an honest traceability claim that is under-tested at the mask level. read the letter →

arxiv 2608.01473 v1 pith:X24HZTKI submitted 2026-08-02 cs.CV cs.CLcs.LG

Guiqiu Liao , Matjaz Jogan , Daniel A. Hashimoto This is my paper
classification cs.CVcs.CLcs.LG
keywords surgicalMLLMslotattentionobject-centricrepresentationvisualtokencompressiongroundedquestionansweringspatialtraceabilityDINOv3featurereconstructionSurg-396K
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper sets out to show that a surgical multimodal large language model does not need hundreds of dense visual tokens to understand a surgical scene. It replaces the dense patch sequence with nine "slots"—region summaries learned by grouping self-supervised DINOv3 visual features—and feeds those slots to the language model as area-labeled tokens. In fast mode, the model answers questions with a 47-token visual prefix instead of the 1,295 tokens used by the leading dense baseline, cutting average total token use by 91.8% while remaining competitive on recognition, grounding, and question-answering benchmarks. In reasoning mode, it also emits explicit area identities, bounding boxes, and masks, so each answer can be traced back to a visual region. This matters because per-frame token cost is the main barrier to real-time, interpretable surgical video understanding.

What carries the argument

The load-bearing mechanism is Slot Attention pretrained by reconstructing DINOv3 patch features. Soft assignment masks from the attention compete across patches for each slot, so each slot becomes a region of the feature map. The same masks are reused after training to softly pool appearance features, compute a centroid and mass for sorting, and provide a literal mask that can be turned into area-ID labels, compressed or uncompressed mask text, and bounding boxes. The bridge from slots to LLM tokens is thus a single representation that simultaneously compresses the visual prefix and retains recoverable spatial support.

What would settle it

Take a surgical frame where a named instrument occupies a small fraction of the image, such as a needle-driver tip, and compute, for the pretrained 9-slot tokenizer, the IoU between each slot's hard mask and the ground-truth instrument mask. If the target's pixels are consistently split across several slots or absorbed into a background slot, so that no single slot reaches IoU near 0.5, then the area-ID supervision is choosing a slot that does not really localize the object, and the claimed traceability is not supported.

Watch

Extended reading notes

Core claim

Slot2Text's central claim is that object-centric compression can serve as the visual interface for a surgical MLLM. A frozen DINOv3 encoder maps the image to patch features; Slot Attention groups those features into nine slot latents with soft assignment masks; a vision-language bridge fuses each slot with its mask-pooled appearance, centroid, and mass and projects it into the LLM input space. The LLM sees a short prefix of area-1: ... area-9: tokens. During training, the model is taught to connect object names to area IDs through grounding supervision that selects the slot with the highest IoU against the target mask and generates mask, box, and location targets. Slot2Text-Fast uses this pr

Load-bearing premise

Slot Attention, trained only to reconstruct DINOv3 patch features, partitions a surgical scene into regions that align with the objects and structures named in the question, so the area-ID picked by the highest-IoU rule in Eq. (5) is a faithful spatial explanation.

Editorial extensions

If this is right

  • Surgical MLLMs can run with a 47-token visual prefix instead of 1,295, cutting per-sample language-model tokens by about 92%, which makes per-frame deployment on surgical video practical.
  • Region-based question answering holds or improves against the dense baseline, because explicit area labels give the LLM a vocabulary for grounded localized questions; unrestricted visual QA remains slightly below the dense baseline.
  • A single model family spans the efficiency-traceability spectrum: Slot2Text-Fast gives cheap direct answers, while Slot2Text-Reason variants emit masks, box coordinates, and area IDs when spatial explanation is required.
  • The visual prefix size is controllable through the slot budget $K$: increasing $K$ from 5 to 13 improves accuracy and mIoU modestly while growing the prefix from 27 to 71 tokens, with diminishing returns after $K=9$.
  • Stage-2 VL-bridge tuning and early uncompressed reasoning traces are required to preserve grounding; removing them drops average F-score, AP@50, and mIoU by 4.92, 3.27, and 6.05 points respectively.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If slot attention on DINOv3 features transfers across the four surgical corpora used for pretraining, it should also transfer to other instrument-dense endoscopic and laparoscopic domains without new image-text pairs, which would make domain-specific surgical assistants cheaper to build.
  • Because slot masks are available at inference, the model could be extended to output segmentations directly through a light mask decoder instead of text-encoded run-length masks, turning the compressed prefix into a genuine panoptic interface.
  • The supplement explicitly cautions that an emitted area-ID is a representational correspondence, not evidence of causal responsibility; testing causal faithfulness would require slot deletion, permutation, or counterfactual mask replacement interventions.
  • The area-ID vocabulary enables a direct audit protocol: ask the model which areas contain a target object, then measure how often the emitted area's mask reaches IoU at least 0.5 with the ground-truth object, turning traceability into a quantitative metric.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. Slot2Text replaces dense patch-token visual inputs in a surgical MLLM with a compact set of region slots. The visual encoder is a frozen DINOv3 ViT; Slot Attention is pretrained to reconstruct DINOv3 patch features (Eq. 2), and each slot is fused with mask-pooled appearance, centroid, and mask mass before projection into the LLM as an 'area-<i>' labeled token. The paper presents two inference modes: Slot2Text-Fast, which answers directly from the compact prefix, and Slot2Text-Reason variants that additionally emit area masks, bounding boxes, and coarse locations via grounding supervision (Eq. 5). Experiments on EndoVis-17/18, CoPESD, and Cholec80-VQA report performance competitive with EndoChat while reducing average total LM tokens from 1,359.15 to 111.10 (91.8%) and the visual prefix from 1,295 to 47 tokens (96.4%). Ablations cover slot budget, grounding supervision, VL-bridge alignment, and fine-tuning schedule.

Significance. If the claims are fully supported, Slot2Text is a meaningful step toward efficient, spatially inspectable visual interfaces for surgical MLLMs. The token-count reduction is directly measured, the method is compared against matched token-compression baselines (Slot-MLLM style, TokenPacker, mask-pooling) on grounding tasks, and the ablation study is careful. The authors also include an honest limitation section, including the caveat that 'traceability' is representational rather than causal. However, two load-bearing parts of the central claim need stronger evidence: (1) the efficiency comparison against EndoChat is potentially confounded by model scale, since EndoChat's backbone size is not reported while Slot2Text uses a Llama 3.2 1B decoder, and (2) the spatial-traceability claim is not validated at the mask level, because the reported grounding metrics are computed on parsed bounding boxes, not on the slot masks that the area-IDs refer to.

major comments (3)
  1. [Table 2 and Figure 3] The efficiency comparison is confounded by LLM scale. Slot2Text uses Llama 3.2 1B, but the EndoChat backbone size is not reported. The reported speed, memory, and token-count gains may therefore be partly attributable to the larger model size of EndoChat rather than to the object-centric visual compression itself. Provide EndoChat's parameter count and a size-matched or FLOP-matched comparison, or explicitly condition the 'much lower cost' claim on model scale. The token-count reduction itself is credible, but the headline 'competitive at much lower cost' is under-supported without controlling for this variable.
  2. [Evaluation Protocol and Direct Answering / Eq. (5)] The spatial-traceability claim is not quantitatively evaluated at the mask level. The evaluation protocol states that grounding metrics (AP@50, mIoU) are computed after predicted coordinates are parsed as bounding boxes; they do not measure the fidelity of the hard slot masks that an emitted area-ID maps back to. Thus, the paper's assertion that Slot2Text-Reason provides 'explicit area identities, locations, and traceable spatial evidence' goes beyond what is measured. I suggest reporting mask-level IoU/AP between the hard mask of the emitted area-ID (or the decoded mask) and the ground-truth object mask, and checking whether the model ever emits invalid or non-existent area-IDs. The qualitative examples in Figures 5 and 8 illustrate the concern: the area attributed to scissors and other instruments includes substantial non-object pixels.
  3. [Equation (5) and Table 1] The area-ID supervision is self-referential in an important way: the target area is selected by argmax IoU between the model's own slot hard mask and the ground-truth object mask. If slot attention fragments an instrument across multiple slots or merges it with background tissue, the supervision reinforces that decomposition rather than correcting it. The external bounding-box metrics partially mitigate this risk, but they do not validate the mask quality that the area-IDs point to. Adding the mask-level evaluation recommended above is necessary to close this gap and to support the claim that area-IDs provide trustworthy spatial correspondence.
minor comments (5)
  1. [Abstract] The phrase 'reducing the average total token consumption by a 91.8%' is grammatically awkward; remove the article 'a'. Also clarify that the reduction is relative to EndoChat, since Table 11 shows TokenPacker uses fewer total LM tokens than Slot2Text-Fast.
  2. [Figure 3] The three-panel figure uses very different scales and its axes are compressed in the reproduction. Please make axis labels and units explicit and readable.
  3. [Table 2] Efficiency columns (LM Tokens, Samples/s) contain dashes for EndoVis rows. State explicitly that efficiency is reported on CoPESD only, or provide the missing measurements.
  4. [Object-Centric Visual Tokenizer] The text says 'Starting from nine initialized queries' but the method should use K initialized queries; nine happens to be the default. Use the general notation to avoid confusion with the ablation.
  5. [Progressive Reasoning Acceleration] The stages are described textually; a small table or algorithm box summarizing the training-data format for each variant (R, R-C, R-L, Fast) would improve reproducibility.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the token-efficiency claim is a measurement, the grounding metrics use external boxes, and the self-referential area-ID targets are training labels, not predictions.

full rationale

The paper's central efficiency claim is a measurement, not a fitted prediction: Table 2 reports actual LM token counts (1,359.15 vs 111.10) and throughput, so the 91.8%/'96.4% reductions are empirical measurements of the constructed architecture. The grounding claims are evaluated externally: AP@50 and mIoU are computed on parsed bounding boxes compared against ground-truth boxes ('Predicted coordinates are parsed as bounding boxes and mapped to the original image space before grounding evaluation'), which does not reduce to the slot masks that area-IDs reference. The one self-referential mechanism is Eq. 5, where the area-ID training target is chosen as the argmax IoU between the model's own hard slot masks and the object mask T. This is a legitimate auxiliary label-generation rule for teaching the LLM to name its own slots, not a claimed prediction of an external truth. The paper's traceability claim is explicitly a representational correspondence, not causal faithfulness: 'We use the term traceability for this representational correspondence; it should not be however interpreted as a guarantee that an area is causally responsible for a generated answer.' The unmeasured fidelity of the emitted slot masks to the named object is a validation gap, not a circular reduction. Author self-citations (Thoracic Robotic Surgery dataset and prior Slot-BERT papers, Liao et al. 2025a,b) are used only for pretraining corpus composition and motivation; the core comparisons are against external baselines on the public Surg-396K benchmark. No derivation step reduces by construction to its own input, and no fitted parameter is renamed as a prediction.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

The method introduces no new physical entities. Slot latents and area-ID text labels are internal model constructs, not independently observable phenomena. The main free parameters are the slot budget and standard training hyperparameters. The self-generated area-ID supervision creates a mild self-referential loop, but external grounding evaluation breaks it.

free parameters (4)
  • Slot budget K = 9
    Chosen by sweeping K in {5, 9, 13} (Figure 4, Tables 7-8); K=9 preferred due to diminishing returns and token cost.
  • Slot attention iterations T = 3
    Standard Slot Attention setting; a design choice, not tuned in this paper.
  • LoRA rank = 128
    Set for the language decoder adaptation; no ablation reported.
  • Slot dropout during Stage 1 = 1 or 2 slots dropped
    Regularization choice in the supplementary implementation details; not analyzed.
assumptions (4)
  • domain assumption DINOv3 patch features are semantically rich enough that reconstructing them with slot attention yields object-relevant regions.
    Invoked in Eq. 2, slot attention pretraining; the paper relies on DINOv3 features transferring to surgical scenes.
  • ad hoc to paper The argmax-IoU criterion in Eq. 5 selects a slot whose mask is a valid spatial target for an object description.
    The grounding labels are generated from the model's own slot assignments; if slots are poorly aligned with objects, the supervision is misleading.
  • domain assumption A 1B Llama 3.2 decoder, adapted by LoRA and a learned bridge, can learn to associate textual area IDs with slot region embeddings.
    Central to both fast answering and grounded reasoning; supported only by the reported experiments.
  • domain assumption The Surg-396K dataset and the authors' 900K-frame pretraining corpus are representative of surgical scenes in the evaluation benchmarks.
    The method's generalization rests on domain overlap between pretraining (laparoscopic, robotic abdominal, thoracic) and evaluation (EndoVis, CoPESD, Cholec80).

how reviews work

0 comments
Cite this review

Pith. "Pith review of Slot2Text: Object-Centric Visual Tokenization for Efficient and Spatially Traceable Surgical MLLMs." pith.science (2026). https://pith.science/paper/X24HZTKI

@misc{pith2026260801473,
  author       = {Pith},
  title        = {Pith review of: Slot2Text: Object-Centric Visual Tokenization for Efficient and Spatially Traceable Surgical MLLMs},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/X24HZTKI}},
  note         = {Machine review of arXiv:2608.01473}
}
read the original abstract

Multimodal large language models (MLLM) for surgical scene understanding typically inject hundreds of dense visual tokens into a language model, leading to costly inference and limited spatial traceability for generated answers. We present Slot2Text, a dual-mode surgical MLLM that replaces dense representations of visual input with a compact set of regions encoded as slot latents. Instead of relying on contrastive alignment of the visual encoder with language, Slot2Text groups self-supervised vision features into a few regions--slots that are consumed by the language model as area-labeled visual tokens. Slot2Text-Fast uses the slot prefix to answer surgical questions. Slot2Text-Reason also identifies and locates areas relevant for reasoning, linking language outputs to corresponding slot tokens, masks or regions. Experiments on multiple visual question answering and visual grounding benchmarks show that Slot2Text-Fast is competitive with state-of-the-art baseline at a much lower cost, reducing the average total token consumption by a 91.8\% and the visual prefix from 1,295 to 47 tokens (a 96.4\% reduction). Slot2Text-Reason trades additional tokens and latency for explicit area identities, locations, and traceable spatial evidence. These results establish compact slot latents as an efficient default visual interface for surgical MLLMs, with grounded reasoning invoked when greater spatial traceability is required.

Figures

Figures reproduced from arXiv: 2608.01473 by the authors.

Figure 1
Figure 1. Comparison of a conventional dense-token MLLM [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Overview of the training framework. (A) Unsupervised Slot Attention pretraining. Frozen visual features are grouped [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 4
Figure 4. Effect of the slot budget K on accuracy, mIoU, and visual-prefix token. EndoVis and CoPESD results are reported separately. In both settings, the slot masks converted from MLLM text output define an explicit area vocabulary, and intermediate responses associate the ultrasound probe, bipolar forceps, and Prograsp forceps with area IDs and normalized bound￾ing boxes. These grounded responses provide spatial context fo… view at source ↗
Figures from the paper (4 more)
Figure 5
Figure 5. Figure 5: Qualitative extended-context reasoning with different slot budgets [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: Qualitative effect of uncompressed-first semantic-descriptor learning. For the same image and query, the reverse [PITH_FULL_IMAGE:figures/full_fig_p015_6.png]
Figure 7
Figure 7. Figure 7: Qualitative localization over one CoPESD scene (three queries) and two EndoVis scenes (five queries). Columns show [PITH_FULL_IMAGE:figures/full_fig_p016_7.png]
Figure 8
Figure 8. Figure 8: Qualitative comparison with EndoChat over two extended-context sessions. In the first example, EndoChat counts [PITH_FULL_IMAGE:figures/full_fig_p017_8.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

255 extracted references · 40 canonical work pages

  1. [1]

    arXiv (Cornell University) , year =

    OpenAI and Achiam, Josh and Adler, Steven and Agarwal, Sandhini and Ahmad, Lama and Akkaya, Ilge and Aleman, Florencia Leoni and Almeida, Diogo and Altenschmidt, Janko and Altman, Sam and Anadkat, Shyamal and Avila, Red and Babuschkin, Igor and Balaji, Suchir and Balcom, Valerie and Baltescu, Paul and Bao, Haiming and Bavarian, Mohammad and Belgum, Jeff a...

  2. [2]

    Medical Image Analysis , publisher =

    Alabi, Oluwatosin and Vercauteren, Tom and Shi, Miaojing , title =. Medical Image Analysis , publisher =. 2025 , volume =. doi:10.1016/j.media.2025.103480 , url =

    work page arXiv 2025

  3. [3]

    arXiv (Cornell University) , year =

    Alayrac, Jean-Baptiste and Jeff Donahue and Pauline Luc and Antoine Miech and Iain Barr and Yana Hasson and Karel Lenc and Arthur Mensch and Katie Millican and Malcolm Reynolds and Roman Ring and Eliza Rutherford and Serkan Cabi and Tengda Han and Zhitao Gong and Sina Samangooei and Marianne Monteiro and Jacob Menick and Sebastian Borgeaud and Andrew Broc...

    arXiv doi:10.48550/arxiv.2204.14198

  4. [4]

    and Kondo, S

    Allan, M. and Kondo, S. and Bodenstedt, S. and Leger, S. and Kadkhodamohammadi, R. and Luengo, I. and Fuentes, F. and Flouty, E. and Mohammed, A. and Pedersen, M., et al. , title =. 2020 , url =

    2020

  5. [5]

    2019 , url =

    Allan, Max and Shvets, Alex and Kurmann, Thomas and Zhang, Zichen and Duggal, Rahul and Su, Yun-Hsuan and Rieke, Nicola and Laina, Iro and Kalavakonda, Niveditha and Bodenstedt, Sebastian and Herrera, Luis and Li, Wenqi and Iglovikov, Vladimir and Luo, Huoling and Yang, Jian and Stoyanov, Danail and Maier-Hein, Lena and Speidel, Stefanie and Azizian, Mahd...

    2019

  6. [6]

    and Sheikh, Mohd Raashid , title =

    Aziz, Hassan and James, Tayler and Remulla, Daphne and Sher, Linda and Genyk, Yuri and Sullivan, Maura E. and Sheikh, Mohd Raashid , title =. Journal of Surgical Education , publisher =. 2021 , volume =. doi:10.1016/j.jsurg.2020.07.037 , url =

    work page doi:10.1016/j.jsurg.2020.07.037 2021

  7. [9]

    2023 IEEE International Conference on Robotics and Automation (ICRA) , publisher =

    Bai, Long and Islam, Mobarakol and Seenivasan, Lalithkumar and Ren, Hongliang , title =. 2023 IEEE International Conference on Robotics and Automation (ICRA) , publisher =. 2023 , pages =. doi:10.1109/icra48891.2023.10160403 , url =

    arXiv 2023

  8. [10]

    Information Fusion , publisher =

    Bai, Long and Wang, Guankun and Islam, Mobarakol and Seenivasan, Lalithkumar and Wang, An and Ren, Hongliang , title =. Information Fusion , publisher =. 2025 , volume =. doi:10.1016/j.inffus.2024.102602 , url =

    work page arXiv 2025

Show all 255 references
  1. [11]

    and Hasan, S.A

    Ben Abacha, A. and Hasan, S.A. and Datla, V.V. and Demner-Fushman, D. and Müller, H. , title =. 2019 , note =

    2019

  2. [12]

    2017 IEEE International Conference on Computer Vision (ICCV) , publisher =

    Ben-younes, Hedi and Cadene, Remi and Cord, Matthieu and Thome, Nicolas , title =. 2017 IEEE International Conference on Computer Vision (ICCV) , publisher =. 2017 , pages =. doi:10.1109/iccv.2017.285 , url =

    2017 doi

  3. [13]

    Proceedings of the AAAI Conference on Artificial Intelligence , publisher =

    Ben-younes, Hedi and Cadene, Remi and Thome, Nicolas and Cord, Matthieu , title =. Proceedings of the AAAI Conference on Artificial Intelligence , publisher =. 2019 , volume =. doi:10.1609/aaai.v33i01.33018102 , url =

    2019 doi

  4. [14]

    Neural Horizons The Uncharted Future of Deep Learning , publisher =

    Malakar, Sourav , title =. Neural Horizons The Uncharted Future of Deep Learning , publisher =. 2026 , pages =. doi:10.65525/svup.9788199778009.2026.224-230 , url =

    2026

  5. [15]

    Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing , publisher =

    Chen, Junying and Gui, Chi and Ouyang, Ruyi and Gao, Anningzhe and Chen, Shunian and Chen, Guiming Hardy and Wang, Xidong and Cai, Zhenyang and Ji, Ke and Wan, Xiang and Wang, Benyou , title =. Proceedings of the 2024 Conference on Empirical Methods in Natural Language Process...

    2024 doi

  6. [16]

    and Zhu, D

    Chen, J. and Zhu, D. and Shen, X. and Li, X. and Liu, Z. and Zhang, P. and Krishnamoor- thi, R. and Chandra, V. and Xiong, Y. and Elhoseiny, M. , title =. 2023 , url =

    2023

  7. [17]

    2023 , url =

    Chen, Jun and Zhu, Deyao and Shen, Xiaoqian and Li, Xiang and Liu, Zechun and Zhang, Pengchuan and Krishnamoorthi, Raghuraman and Chandra, Vikas and Xiong, Yunyang and Elhoseiny, Mohamed , title =. 2023 , url =

    2023

  8. [18]

    2024 IEEE International Conference on Robotics and Automation (ICRA) , publisher =

    Chen, Kexin and Du, Yuyang and You, Tao and Islam, Mobarakol and Guo, Ziyu and Jin, Yueming and Chen, Guangyong and Heng, Pheng-Ann , title =. 2024 IEEE International Conference on Robotics and Automation (ICRA) , publisher =. 2024 , pages =. doi:10.1109/icra57147.2024.1061060...

    2024

  9. [19]

    Advances in Neural Information Processing Systems 37 , publisher =

    Chen, Pengcheng and Ye, Jin and Wang, Guoan and Li, Yanjun and Deng, Zhongying and Li, Wei and Li, Tianbin and Duan, Haodong and Huang, Ziyan and Su, Yanzhou and Wang, Benyou and Zhang, Shaoting and Fu, Bin and Cai, Jianfei and Zhuang, Bohan and Seibel, Eric and Qiao, Yu and H...

    2024 doi

  10. [20]

    Surgical Endoscopy , publisher =

    Chen, Richard and Rodrigues Armijo, Priscila and Krause, Crystal and Siu, Ka-Chun and Oleynikov, Dmitry , title =. Surgical Endoscopy , publisher =. 2020 , volume =. doi:10.1007/s00464-019-06775-1 , url =

    2020 doi

  11. [21]

    Science China Information Sciences , publisher =

    Chen, Zhe and Wang, Weiyun and Tian, Hao and Ye, Shenglong and Gao, Zhangwei and Cui, Erfei and Tong, Wenwen and Hu, Kongzhi and Luo, Jiapeng and Ma, Zheng and Ma, Ji and Wang, Jiaqi and Dong, Xiaoyi and Yan, Hang and Guo, Hewei and He, Conghui and Shi, Botian and Jin, Zhenjia...

    2024

  12. [22]

    Proceedings of the 1st International Conference on Data Science and Engineering , publisher =

    Liu, Boyu , title =. Proceedings of the 1st International Conference on Data Science and Engineering , publisher =. 2024 , pages =. doi:10.5220/0012829800004547 , url =

    2024 doi

  13. [23]

    2023 IEEE International Conference on Medical Artificial Intelligence (MedAI) , publisher =

    Dou, Yutao and Zhao, Xiongjun and Zou, Haitao and Xiao, Jian and Xi, Peng and Peng, Shaoliang , title =. 2023 IEEE International Conference on Medical Artificial Intelligence (MedAI) , publisher =. 2023 , pages =. doi:10.1109/medai59581.2023.00017 , url =

    2023

  14. [24]

    and Ligero, Marta and Sainath, Srividhya and Ghaffari Laleh, Narmin and El Nahhas, Omar S

    Ferber, Dyke and Wölflein, Georg and Wiest, Isabella C. and Ligero, Marta and Sainath, Srividhya and Ghaffari Laleh, Narmin and El Nahhas, Omar S. M. and Müller-Franzes, Gustav and Jäger, Dirk and Truhn, Daniel and Kather, Jakob Nikolas , title =. Nature Communications , publi...

    2024 doi

  15. [25]

    npj Digital Medicine , publisher =

    Goetz, Lea and Seedat, Nabeel and Vandersluis, Robert and van der Schaar, Mihaela , title =. npj Digital Medicine , publisher =. 2024 , volume =. doi:10.1038/s41746-024-01127-3 , url =

    2024 doi

  16. [26]

    and Garrido-Merchan, E.C

    Gozalo-Brizuela, R. and Garrido-Merchan, E.C. , title =. 2023 , url =

    2023

  17. [27]

    2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition , publisher =

    Li, Yikang and Duan, Nan and Zhou, Bolei and Chu, Xiao and Ouyang, Wanli and Wang, Xiaogang and Zhou, Ming , title =. 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition , publisher =. 2018 , pages =. doi:10.1109/cvpr.2018.00640 , url =

    2018

  18. [28]

    and Xu, M

    He, R. and Xu, M. and Das, A. and Khan, D.Z. and Bano, S. and Marcus, H.J. and Stoyanov, D. and Clarkson, M.J. and Islam, M. , title =. 2024 , note =

    2024

  19. [29]

    and Nie, Y

    He, S. and Nie, Y. and Chen, Z. and Cai, Z. and Wang, H. and Yang, S. and Chen, H. , title =. 2024 , url =

    2024

  20. [30]

    and Cheng, Y

    Hou, W. and Cheng, Y. and Xu, K. and Hu, Y. and Li, W. and Liu, J. , title =. 2024 , url =

    2024

  21. [31]

    and Shen, Y

    Hu, E.J. and Shen, Y. and Wallis, P. and Allen-Zhu, Z. and Li, Y. and Wang, S. and Wang, L. and Chen, W. , title =. 2021 , url =

    2021

  22. [33]

    and Li, C

    Huang, W. and Li, C. and Yang, H. and Liu, J. and Liang, Y. and Zheng, H. and Wang, S. , title =. 2024 , note =

    2024

  23. [34]

    Medical Image Analysis , publisher =

    Islam, Mobarakol and VS, Vibashan and Lim, Chwee Ming and Ren, Hongliang , title =. Medical Image Analysis , publisher =. 2021 , volume =. doi:10.1016/j.media.2020.101837 , url =

    2021

  24. [35]

    and Jeong, C.W

    Jin, J. and Jeong, C.W. , title =. 2024 , url =

    2024

  25. [36]

    and Hussain, M

    Khanam, R. and Hussain, M. , title =. 2024 , url =

    2024

  26. [37]

    ACM Computing Surveys , publisher =

    Kuang, Jiayi and Shen, Ying and Xie, Jingyou and Luo, Haohao and Xu, Zhe and Li, Ronghao and Li, Yinghui and Cheng, Xianfeng and Lin, Xika and Han, Yu , title =. ACM Computing Surveys , publisher =. 2025 , volume =. doi:10.1145/3711680 , url =

    2025 doi

  27. [38]

    , title =

    Bigolin Lanfredi, Ricardo and Mukherjee, Pritam and Summers, Ronald M. , title =. Medical Image Analysis , publisher =. 2025 , volume =. doi:10.1016/j.media.2024.103383 , url =

    2025

  28. [40]

    Advances in Neural Information Processing Systems 36 , publisher =

    Li, Chunyuan and Wong, Cliff and Zhang, Sheng and Usuyama, Naoto and Liu, Haotian and Yang, Jianwei and Naumann, Tristan and Poon, Hoifung and Gao, Jianfeng , title =. Advances in Neural Information Processing Systems 36 , publisher =. 2023 , pages =. doi:10.52202/075280-1240 , url =

    2023 doi

  29. [41]

    and Li, D

    Li, J. and Li, D. and Savarese, S. and Hoi, S. , title =. 2023 , note =

    2023

  30. [42]

    and Skinner, G

    Li, J. and Skinner, G. and Yang, G. and Quaranto, B.R. and Schwaitzberg, S.D. and Kim, P.C. and Xiong, J. , title =. 2024 , url =

    2024

  31. [43]

    2019 , url =

    Li, Liunian Harold and Yatskar, Mark and Yin, Da and Hsieh, Cho-Jui and Chang, Kai-Wei , title =. 2019 , url =

    2019

  32. [44]

    Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) , publisher =

    Li, Xiang Lisa and Holtzman, Ari and Fried, Daniel and Liang, Percy and Eisner, Jason and Hashimoto, Tatsunori and Zettlemoyer, Luke and Lewis, Mike , title =. Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) , pub...

    2023 doi

  33. [45]

    Computer Vision -- ECCV 2024 , series =

    Lin, Ziyi and Liu, Dongyang and Zhang, Renrui and Gao, Peng and Qiu, Longtian and Xiao, Han and Qiu, Han and Shao, Wenqi and Chen, Keqin and Han, Jiaming and Huang, Siyuan and Zhang, Yichi and He, Xuming and Qiao, Yu and Li, Hongsheng , title =. Computer Vision -- ECCV 2024 , ...

    2024 doi

  34. [46]

    2026 , doi =

    Abbasi, Ahmed and Haghighat, Mohammadreza and Naseem, Usman and Rahimi Azghadi, Mostafa , title =. 2026 , doi =

    2026

  35. [47]

    and Zhu, T

    Liu, F. and Zhu, T. and Wu, X. and Yang, B. and You, C. and Wang, C. and Lu, L. and Liu, Z. and Zheng, Y. and Sun, X., et al. , title =. 2023 , note =

    2023

  36. [49]

    Advances in Neural Information Processing Systems 36 , publisher =

    Liu, Haotian and Li, Chunyuan and Wu, Qingyang and Lee, Yong Jae , title =. Advances in Neural Information Processing Systems 36 , publisher =. 2023 , pages =. doi:10.52202/075280-1516 , url =

    2023 doi

  37. [50]

    and Hutter, F

    Loshchilov, I. and Hutter, F. , title =. 2017 , url =

    2017

  38. [51]

    Medical Image Analysis , publisher =

    Lu, Yinbin and Wang, Alan , title =. Medical Image Analysis , publisher =. 2025 , volume =. doi:10.1016/j.media.2025.103514 , url =

    2025

  39. [52]

    IEEE Transactions on Biomedical Engineering , publisher =

    Mariani, Andrea and Pellegrini, Edoardo and De Momi, Elena , title =. IEEE Transactions on Biomedical Engineering , publisher =. 2021 , volume =. doi:10.1109/tbme.2020.3011867 , url =

    2021

  40. [53]

    and Alapatt, D

    Nwoye, C.I. and Alapatt, D. and Yu, T. and Vardazaryan, A. and Xia, F. and Zhao, Z. and Xia, T. and Jia, F. and Yang, Y. and Wang, H., et al. , title =. 2023 , note =

    2023

  41. [54]

    Medical Image Analysis , publisher =

    Nwoye, Chinedu Innocent and Padoy, Nicolas , title =. Medical Image Analysis , publisher =. 2025 , volume =. doi:10.1016/j.media.2024.103438 , url =

    2025

  42. [55]

    and Darcet, T

    Oquab, M. and Darcet, T. and Moutakanni, T. and Vo, H. and Szafraniec, M. and Khali- dov, V. and Fernandez, P. and Haziza, D. and Massa, F. and El-Nouby, A., et al. , title =. 2023 , url =

    2023

  43. [56]

    Lecture Notes in Computer Science , publisher =

    Pellegrini, Chantal and Keicher, Matthias and Özsoy, Ege and Navab, Nassir , title =. Lecture Notes in Computer Science , publisher =. 2023 , pages =. doi:10.1007/978-3-031-43904-9\_40 , url =

    2023 doi

  44. [57]

    2024 International Joint Conference on Neural Networks (IJCNN) , publisher =

    Peng, Peixi and Fan, Wanshu and Liu, Wenfei and Yang, Xing and Zhou, Dongsheng , title =. 2024 International Joint Conference on Neural Networks (IJCNN) , publisher =. 2024 , pages =. doi:10.1109/ijcnn60899.2024.10650493 , url =

    2024

  45. [58]

    and Kim, J.W

    Radford, A. and Kim, J.W. and Hallacy, C. and Ramesh, A. and Goh, G. and Agarwal, S. and Sastry, G. and Askell, A. and Mishkin, P. and Clark, J., et al. , title =. 2021 , note =

    2021

  46. [59]

    and Tu, T

    Saab, K. and Tu, T. and Weng, W.H. and Tanno, R. and Stutz, D. and Wulczyn, E. and Zhang, F. and Strother, T. and Park, C. and Vedadi, E., et al. , title =. 2024 , url =

    2024

  47. [60]

    and Cho, J

    Schmidgall, S. and Cho, J. and Zakka, C. and Hiesinger, W. , title =. 2024 , url =

    2024

  48. [61]

    Lecture Notes in Computer Science , publisher =

    Seenivasan, Lalithkumar and Islam, Mobarakol and Kannan, Gokul and Ren, Hongliang , title =. Lecture Notes in Computer Science , publisher =. 2023 , pages =. doi:10.1007/978-3-031-43996-4\_27 , url =

    2023 doi

  49. [62]

    Lecture Notes in Computer Science , publisher =

    Seenivasan, Lalithkumar and Islam, Mobarakol and Krishna, Adithya K and Ren, Hongliang , title =. Lecture Notes in Computer Science , publisher =. 2022 , pages =. doi:10.1007/978-3-031-16449-1\_4 , url =

    2022 doi

  50. [63]

    , title =

    Sharma, Dhruv and Purushotham, Sanjay and Reddy, Chandan K. , title =. Scientific Reports , publisher =. 2021 , volume =. doi:10.1038/s41598-021-98390-1 , url =

    2021 doi

  51. [64]

    and Kida, T

    Shibata, Y. and Kida, T. and Fukamachi, S. and Takeda, M. and Shinohara, A. and Shi- nohara, T. and Arikawa, S. , title =. 1999 , note =

    1999

  52. [65]

    and Liu, Jiangfeng and Wang, Jing and Wang, Binglu and Yang, Juntao , title =

    Song, Meiyue and Wang, Jiarui and Yu, Zhihua and Wang, Jiaxin and Yang, Le and Lu, Yuting and Li, Baicun and Wang, Xue and Wang, Xiaoxu and Huang, Qinghua and Li, Zhijun and Kanellakis, Nikolaos I. and Liu, Jiangfeng and Wang, Jing and Wang, Binglu and Yang, Juntao , title =. ...

    2024

  53. [66]

    Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing , publisher =

    Song, Xinying and Salcianu, Alex and Song, Yang and Dopson, Dave and Zhou, Denny , title =. Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing , publisher =. 2021 , pages =. doi:10.18653/v1/2021.emnlp-main.160 , url =

    2021 doi

  54. [67]

    and Mullappilly, Sahal Shaji and Cholakkal, Hisham and Anwer, Rao Muhammad and Khan, Salman and Laaksonen, Jorma and Khan, Fahad , title =

    Thawakar, Omkar Chakradhar and Shaker, Abdelrahman M. and Mullappilly, Sahal Shaji and Cholakkal, Hisham and Anwer, Rao Muhammad and Khan, Salman and Laaksonen, Jorma and Khan, Fahad , title =. Proceedings of the 23rd Workshop on Biomedical Natural Language Processing , publis...

    2024 doi

  55. [68]

    Proceedings of the 38th International Conference on Machine Learning , publisher =

    Touvron, Hugo and Cord, Matthieu and Douze, Matthijs and Massa, Francisco and Sablayrolles, Alexandre and Jégou, Hervé , title =. Proceedings of the 38th International Conference on Machine Learning , publisher =. 2021 , volume =

    2021

  56. [69]

    and Martin, L

    Touvron, H. and Martin, L. and Stone, K. and Albert, P. and Almahairi, A. and Babaei, Y. and Bashlykov, N. and Batra, S. and Bhargava, P. and Bhosale, S., et al. , title =. 2023 , url =

    2023

  57. [70]

    and Azizi, S

    Tu, T. and Azizi, S. and Driess, D. and Schaekermann, M. and Amin, M. and Chang, P.C. and Carroll, A. and Lau, C. and Tanno, R. and Ktena, I., et al. , title =. 2024 , note =

    2024

  58. [71]

    and Shehata, Sherif and Mutter, Didier and Marescaux, Jacques and de Mathelin, Michel and Padoy, Nicolas , title =

    Twinanda, Andru P. and Shehata, Sherif and Mutter, Didier and Marescaux, Jacques and de Mathelin, Michel and Padoy, Nicolas , title =. IEEE Transactions on Medical Imaging , publisher =. 2017 , volume =. doi:10.1109/tmi.2016.2593957 , url =

    2017

  59. [72]

    From Human Attention to Computational Attention , publisher =

    Mineault, Patrick , title =. From Human Attention to Computational Attention , publisher =. 2025 , pages =. doi:10.1007/978-3-031-84300-6\_13 , url =

    2025 doi

  60. [73]

    Wagner, Martin and Müller-Stich, Beat-Peter and Kisilenko, Anna and Tran, Duc and Heger, Patrick and Mündermann, Lars and Lubotsky, David M and Müller, Benjamin and Davitashvili, Tornike and Capek, Manuela and Reinke, Annika and Reid, Carissa and Yu, Tong and Vardazaryan, Armi...

    2023

  61. [74]

    2024 , url =

    Wang, Guankun and Bai, Long and Nah, Wan Jun and Wang, Jie and Zhang, Zhaoxi and Chen, Zhen and Wu, Jinlin and Islam, Mobarakol and Liu, Hongbin and Ren, Hongliang , title =. 2024 , url =

    2024

  62. [75]

    2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , publisher =

    Wang, Guanqun and Liu, Jiaming and Li, Chenxuan and Zhang, Yuan and Ma, Junpeng and Wei, Xinyu and Zhang, Kevin and Chong, Maurice and Zhang, Renrui and Liu, Yijiang and Zhang, Shanghang , title =. 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , pu...

    2024

  63. [76]

    Proceedings of the 33rd ACM International Conference on Multimedia , publisher =

    Wang, Guankun and Xiao, Han and Zhang, Renrui and Gao, Huxin and Bai, Long and Yang, Xiaoxiao and Li, Zhen and Li, Hongsheng and Ren, Hongliang , title =. Proceedings of the 33rd ACM International Conference on Multimedia , publisher =. 2025 , pages =. doi:10.1145/3746027.3758...

    2025

  64. [77]

    2024 , doi =

    Zhao, Xiujie and Gao, Yu and Zhang, Yimeng , title =. 2024 , doi =

    2024

  65. [78]

    2024 , url =

    Wang, Peng and Bai, Shuai and Tan, Sinan and Wang, Shijie and Fan, Zhihao and Bai, Jinze and Chen, Keqin and Liu, Xuejing and Wang, Jialin and Ge, Wenbin and Fan, Yang and Dang, Kai and Du, Mengfei and Ren, Xuancheng and Men, Rui and Liu, Dayiheng and Zhou, Chang and Zhou, Jin...

    2024

  66. [79]

    and Duan, Y

    Wang, R. and Duan, Y. and Li, J. and Pang, P. and Tan, T. , title =. 2023 , url =

    2023

  67. [80]

    Communications Engineering , publisher =

    Wang, Sheng and Zhao, Zihao and Ouyang, Xi and Liu, Tianming and Wang, Qian and Shen, Dinggang , title =. Communications Engineering , publisher =. 2024 , volume =. doi:10.1038/s44172-024-00271-8 , url =

    2024 doi

  68. [81]

    and Shu, M

    Xue, L. and Shu, M. and Awadalla, A. and Wang, J. and Yan, A. and Purushwalkam, S. and Zhou, H. and Prabhu, V. and Dai, Y. and Ryoo, M.S., et al. , title =. 2024 , url =

    2024

  69. [82]

    and Yu, B

    Yang, A. and Yu, B. and Li, C. and Liu, D. and Huang, F. and Huang, H. and Jiang, J. and Tu, J. and Zhang, J. and Zhou, J., et al. , title =. 2025 , url =

    2025

  70. [83]

    and Yu, T

    Yao, Y. and Yu, T. and Zhang, A. and Wang, C. and Cui, J. and Zhu, H. and Cai, T. and Li, H. and Zhao, W. and He, Z., et al. , title =. 2024 , url =

    2024

  71. [84]

    IEEE Transactions on Neural Networks and Learning Systems , publisher =

    Yu, Zhou and Yu, Jun and Xiang, Chenchao and Fan, Jianping and Tao, Dacheng , title =. IEEE Transactions on Neural Networks and Learning Systems , publisher =. 2018 , volume =. doi:10.1109/tnnls.2018.2817340 , url =

    2018

  72. [85]

    and Navab, Nassir and Srivastav, Vinkle and Padoy, Nicolas , title =

    Yuan, Kun and Kattel, Manasi and Lavanchy, Joël L. and Navab, Nassir and Srivastav, Vinkle and Padoy, Nicolas , title =. International Journal of Computer Assisted Radiology and Surgery , publisher =. 2024 , volume =. doi:10.1007/s11548-024-03141-y , url =

    2024 doi

  73. [86]

    Advances in Neural Information Processing Systems 37 , publisher =

    Yuan, Kun and Srivastav, Vinkle and Navab, Nassir and Padoy, Nicolas , title =. Advances in Neural Information Processing Systems 37 , publisher =. 2024 , pages =. doi:10.52202/079017-3907 , url =

    2024 doi

  74. [87]

    Lecture Notes in Computer Science , publisher =

    Yuan, Kun and Srivastav, Vinkle and Navab, Nassir and Padoy, Nicolas , title =. Lecture Notes in Computer Science , publisher =. 2024 , pages =. doi:10.1007/978-3-031-72089-5\_29 , url =

    2024 doi

  75. [88]

    and Marescaux, Jacques and Mascagni, Pietro and Navab, Nassir and Padoy, Nicolas , title =

    Yuan, Kun and Srivastav, Vinkle and Yu, Tong and Lavanchy, Joël L. and Marescaux, Jacques and Mascagni, Pietro and Navab, Nassir and Padoy, Nicolas , title =. Medical Image Analysis , publisher =. 2025 , volume =. doi:10.1016/j.media.2025.103644 , url =

    2025

  76. [89]

    Zhang, Kai and Zhou, Rong and Adhikarla, Eashan and Yan, Zhiling and Liu, Yixin and Yu, Jun and Liu, Zhengliang and Chen, Xun and Davison, Brian D. and Ren, Hui and Huang, Jing and Chen, Chen and Zhou, Yuyin and Fu, Sunyang and Liu, Wei and Liu, Tianming and Li, Xiang and Chen...

    2023

  77. [90]

    Visual Computing for Industry, Biomedicine, and Art , publisher =

    Zhang, Yue and Fan, Wanshu and Peng, Peixi and Yang, Xin and Zhou, Dongsheng and Wei, Xiaopeng , title =. Visual Computing for Industry, Biomedicine, and Art , publisher =. 2024 , volume =. doi:10.1186/s42492-024-00160-z , url =

    2024 doi

  78. [91]

    2023 , doi =

    Zhou, Juexiao and He, Xiaonan and Sun, Liyuan and Xu, Jiannan and Chen, Xiuying and Chu, Yuetan and Zhou, Longxi and Liao, Xingyu and Zhang, Bin and Gao, Xin , title =. 2023 , doi =

    2023

  79. [92]

    and Bai, L

    Zhou, Y. and Bai, L. and Cai, S. and Deng, B. and Xu, X. and Shen, H.T. , title =. 2025 , note =

    2025

  80. [93]

    Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) , publisher =

    Zhou, Yucheng and Song, Lingran and Shen, Jianbing , title =. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) , publisher =. 2025 , pages =. doi:10.18653/v1/2025.acl-long.636 , url =

    2025 doi

  81. [94]

    2024 , doi =

    Tao, Shiquan and Wongsawat, Arthit , title =. 2024 , doi =

    2024

  82. [95]

    Proceedings of the Language Resources and Evaluation Conference , publisher =

    Zhu, Zhihong and Zhang, Yunyan and Cheng, Xuxin and Huang, Zhiqi and Xu, Derong and Wu, Xian and Zheng, Yefeng , title =. Proceedings of the Language Resources and Evaluation Conference , publisher =. 2024 , pages =. doi:10.63317/4jh5knz54ake , url =

    2024 doi

  83. [96]

    Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , year =

    Yu, Zhou and Yu, Jun and Cui, Yuhao and Tao, Dacheng and Tian, Qi , title =. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , year =. doi:10.1109/CVPR.2019.00644 , url =

    2019

  84. [97]

    International Journal of Computer Vision , volume =

    TokenPacker: Efficient Visual Projector for Multimodal LLM , author =. International Journal of Computer Vision , volume =. 2025 , publisher =

    2025

  85. [98]

    Proceedings of Machine learning and systems , volume=

    Federated optimization in heterogeneous networks , author=. Proceedings of Machine learning and systems , volume=

  86. [99]

    Lan, Mengcheng and Chen, Chaofeng and Zhou, Yue and Xu, Jiaxing and Ke, Yiping and Wang, Xinjiang and Feng, Litong and Zhang, Wei , booktitle=

  87. [100]

    arXiv preprint arXiv:2508.10104 , year=

    DINOv3 , author=. arXiv preprint arXiv:2508.10104 , year=

    arXiv

  88. [101]

    arXiv preprint arXiv:2501.12477 , year=

    Slot-BERT: Self-supervised Object Discovery in Surgical Video , author=. arXiv preprint arXiv:2501.12477 , year=

  89. [102]

    International Conference on Medical Image Computing and Computer-Assisted Intervention , pages=

    Future slot prediction for unsupervised object discovery in surgical video , author=. International Conference on Medical Image Computing and Computer-Assisted Intervention , pages=. 2025 , organization=

    2025

  90. [103]

    arXiv preprint arXiv:2403.12945 , year=

    Droid: A large-scale in-the-wild robot manipulation dataset , author=. arXiv preprint arXiv:2403.12945 , year=

    arXiv

  91. [104]

    Proceedings of the Computer Vision and Pattern Recognition Conference , pages=

    Ctrl-o: language-controllable object-centric visual representation learning , author=. Proceedings of the Computer Vision and Pattern Recognition Conference , pages=

  92. [105]

    Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

    Spot: Self-training with patch-order permutation for object-centric learning with autoregressive transformers , author=. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

  93. [106]

    Proceedings of the Computer Vision and Pattern Recognition Conference , pages=

    Temporally consistent object-centric learning by contrasting slots , author=. Proceedings of the Computer Vision and Pattern Recognition Conference , pages=

  94. [107]

    5: Improved baselines for agglomerative vision foundation models , author=

    Radiov2. 5: Improved baselines for agglomerative vision foundation models , author=. Proceedings of the Computer Vision and Pattern Recognition Conference , pages=

  95. [108]

    International conference on machine learning , pages=

    Multi-object representation learning with iterative variational inference , author=. International conference on machine learning , pages=. 2019 , organization=

    2019

  96. [109]

    International Journal of Computer Vision , volume=

    Universal representations: A unified look at multiple task and domain learning , author=. International Journal of Computer Vision , volume=. 2024 , publisher=

    2024

  97. [110]

    IEEE Transactions on Circuits and Systems for Video Technology , year=

    FedDiff: Diffusion model driven federated learning for multi-modal and multi-clients , author=. IEEE Transactions on Circuits and Systems for Video Technology , year=

  98. [111]

    Proceedings of the IEEE/CVF international conference on computer vision , pages=

    Segment anything , author=. Proceedings of the IEEE/CVF international conference on computer vision , pages=

  99. [112]

    Proceedings of the IEEE conference on computer vision and pattern recognition , pages=

    Squeeze-and-excitation networks , author=. Proceedings of the IEEE conference on computer vision and pattern recognition , pages=

  100. [113]

    Journal of Machine Learning Research , volume =

    Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity , author =. Journal of Machine Learning Research , volume =

  101. [114]

    International Conference on Learning Representations (ICLR) , year =

    Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer , author =. International Conference on Learning Representations (ICLR) , year =

  102. [115]

    Advances in neural information processing systems , volume=

    Learning multiple visual domains with residual adapters , author=. Advances in neural information processing systems , volume=

  103. [116]

    arXiv preprint arXiv:1804.03235 , year=

    Large scale distributed neural network training through online distillation , author=. arXiv preprint arXiv:1804.03235 , year=

    arXiv

  104. [117]

    SIAM journal on control and optimization , volume=

    Acceleration of stochastic approximation by averaging , author=. SIAM journal on control and optimization , volume=. 1992 , publisher=

    1992

  105. [118]

    Nature communications , volume=

    Communication-efficient federated learning via knowledge distillation , author=. Nature communications , volume=. 2022 , publisher=

    2022

  106. [119]

    International conference on machine learning , pages=

    Data-free knowledge distillation for heterogeneous federated learning , author=. International conference on machine learning , pages=. 2021 , organization=

    2021

  107. [120]

    Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

    Region-based representations revisited , author=. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

  108. [121]

    How to grow a mind: Statistics, structure, and abstraction , volume =

    Tenenbaum, Joshua B and Kemp, Charles and Griffiths, Thomas L and Goodman, Noah D , journal =. How to grow a mind: Statistics, structure, and abstraction , volume =

  109. [122]

    Artificial intelligence and statistics , pages=

    Communication-efficient learning of deep networks from decentralized data , author=. Artificial intelligence and statistics , pages=. 2017 , organization=

    2017

  110. [123]

    Conference on Machine Learning and Systems (MLSys) , year=

    FedProx: Mitigating System Heterogeneity in Federated Learning , author=. Conference on Machine Learning and Systems (MLSys) , year=

  111. [124]

    arXiv preprint arXiv:1610.05492 , year=

    Federated Learning: Strategies for Improving Communication Efficiency , author=. arXiv preprint arXiv:1610.05492 , year=

    arXiv

  112. [125]

    Advances in Neural Information Processing Systems (NeurIPS) , year=

    SlotFormer: Unsupervised Dynamic Scene Decomposition , author=. Advances in Neural Information Processing Systems (NeurIPS) , year=

  113. [126]

    The Eleventh International Conference on Learning Representations , year=

    Improving Object-Centric Learning with Query Optimization , author=. The Eleventh International Conference on Learning Representations , year=

  114. [127]

    Conference on Empirical Methods in Natural Language Processing (EMNLP) , year=

    Federated CLIP: Harnessing Decentralized Vision-Language Models , author=. Conference on Empirical Methods in Natural Language Processing (EMNLP) , year=

  115. [128]

    IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , year=

    Federated Foundation Models: Privacy-Preserving Adaptation , author=. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , year=

  116. [129]

    Advances in Neural Information Processing Systems (NeurIPS) , year=

    Learning Multiple Visual Domains with Residual Adapters , author=. Advances in Neural Information Processing Systems (NeurIPS) , year=

  117. [130]

    IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI) , year=

    Multi-Task Learning for Dense Prediction Tasks: A Survey , author=. IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI) , year=

  118. [131]

    International Conference on Learning Representations (ICLR) , year=

    FedAdapter: Efficient Federated Learning with Local Adapters , author=. International Conference on Learning Representations (ICLR) , year=

  119. [132]

    The reviewing of object files: Object-specific integration of information , volume =

    Kahneman, Daniel and Treisman, Anne and Gibbs, Brian J , journal =. The reviewing of object files: Object-specific integration of information , volume =

  120. [133]

    Monet: Unsupervised scene decomposition and representation , volume =

    Burgess, Christopher P and Matthey, Loic and Watters, Nicholas and Kabra, Rishabh and Higgins, Irina and Botvinick, Matt and Lerchner, Alexander , journal =. Monet: Unsupervised scene decomposition and representation , volume =

  121. [134]

    Multi-Object Representation Learning with Iterative Variational Inference , volume =

    Klaus Greff and Rapha. Multi-Object Representation Learning with Iterative Variational Inference , volume =. Proceedings of the 36th International Conference on Machine Learning,

  122. [135]

    Object-Centric Learning with Slot Attention , year =

    Francesco Locatello and Dirk Weissenborn and Thomas Unterthiner and Aravindh Mahendran and Georg Heigold and Jakob Uszkoreit and Alexey Dosovitskiy and Thomas Kipf , booktitle =. Object-Centric Learning with Slot Attention , year =

  123. [136]

    ArXiv preprint , title =

    Greff, Klaus and Van Steenkiste, Sjoerd and Schmidhuber, J. ArXiv preprint , title =

  124. [137]

    Bridging the Gap to Real-World Object-Centric Learning , year =

    Maximilian Seitzer and Max Horn and Andrii Zadaianchuk and Dominik Zietlow and Tianjun Xiao and Carl. Bridging the Gap to Real-World Object-Centric Learning , year =. The Eleventh International Conference on Learning Representations,

  125. [138]

    Adaptive slot attention: Object discovery with dynamic slot number , year =

    Fan, Ke and Bai, Zechen and Xiao, Tianjun and He, Tong and Horn, Max and Fu, Yanwei and Locatello, Francesco and Zhang, Zheng , booktitle =. Adaptive slot attention: Object discovery with dynamic slot number , year =

  126. [139]

    Object-centric architectures enable efficient causal representation learning , volume =

    Mansouri, Amin and Hartford, Jason and Zhang, Yan and Bengio, Yoshua , journal =. Object-centric architectures enable efficient causal representation learning , volume =

  127. [140]

    Object-Centric Slot Diffusion , year =

    Jindong Jiang and Fei Deng and Gautam Singh and Sungjin Ahn , booktitle =. Object-Centric Slot Diffusion , year =

  128. [141]

    SlotDiffusion: Object-Centric Generative Modeling with Diffusion Models , year =

    Ziyi Wu and Jingyu Hu and Wuyue Lu and Igor Gilitschenski and Animesh Garg , booktitle =. SlotDiffusion: Object-Centric Generative Modeling with Diffusion Models , year =

  129. [142]

    Unmasking the inductive biases of unsupervised object representations for video sequences , volume =

    Weis, Marissa A and Chitta, Kashyap and Sharma, Yash and Brendel, Wieland and Bethge, Matthias and Geiger, Andreas and Ecker, Alexander S , journal =. Unmasking the inductive biases of unsupervised object representations for video sequences , volume =

  130. [143]

    Conditional Object-Centric Learning from Video , year =

    Thomas Kipf and Gamaleldin Fathy Elsayed and Aravindh Mahendran and Austin Stone and Sara Sabour and Georg Heigold and Rico Jonschkowski and Alexey Dosovitskiy and Klaus Greff , booktitle =. Conditional Object-Centric Learning from Video , year =

  131. [144]

    Self-supervised Object-Centric Learning for Videos , year =

    G. Self-supervised Object-Centric Learning for Videos , year =. Advances in Neural Information Processing Systems 36: Annual Conference on Neural Information Processing Systems 2023, NeurIPS 2023, New Orleans, LA, USA, December 10 - 16, 2023 , timestamp =

    2023

  132. [145]

    Object-Centric Learning for Real-World Videos by Predicting Temporal Feature Similarities , year =

    Andrii Zadaianchuk and Maximilian Seitzer and Georg Martius , booktitle =. Object-Centric Learning for Real-World Videos by Predicting Temporal Feature Similarities , year =

  133. [146]

    SlotFormer: Unsupervised Visual Dynamics Simulation with Object-Centric Models , year =

    Ziyi Wu and Nikita Dvornik and Klaus Greff and Thomas Kipf and Animesh Garg , booktitle =. SlotFormer: Unsupervised Visual Dynamics Simulation with Object-Centric Models , year =

  134. [147]

    Ondrej Biza and Sjoerd van Steenkiste and Mehdi S. M. Sajjadi and Gamaleldin Fathy Elsayed and Aravindh Mahendran and Thomas Kipf , booktitle =. Invariant Slot Attention: Object Discovery with Slot-Centric Reference Frames , volume =

  135. [148]

    Guided Slot Attention for Unsupervised Video Object Segmentation , year =

    Lee, Minhyeok and Cho, Suhwan and Lee, Dogyoon and Park, Chaewon and Lee, Jungho and Lee, Sangyoun , booktitle =. Guided Slot Attention for Unsupervised Video Object Segmentation , year =

  136. [149]

    Semantics Meets Temporal Correspondence: Self-supervised Object-centric Learning in Videos , year =

    Rui Qian and Shuangrui Ding and Xian Liu and Dahua Lin , booktitle =. Semantics Meets Temporal Correspondence: Self-supervised Object-centric Learning in Videos , year =

  137. [150]

    Discovering Objects that Can Move , year =

    Zhipeng Bao and Pavel Tokmakov and Allan Jabri and Yu. Discovering Objects that Can Move , year =

  138. [151]

    Object Discovery from Motion-Guided Tokens , year =

    Zhipeng Bao and Pavel Tokmakov and Yu. Object Discovery from Motion-Guided Tokens , year =

  139. [152]

    Simple Unsupervised Object-Centric Learning for Complex and Naturalistic Videos , year =

    Gautam Singh and Yi. Simple Unsupervised Object-Centric Learning for Complex and Naturalistic Videos , year =. Advances in Neural Information Processing Systems 35: Annual Conference on Neural Information Processing Systems 2022, NeurIPS 2022, New Orleans, LA, USA, November 28...

    2022

  140. [153]

    Parallelized Spatiotemporal Slot Binding for Videos , year =

    Singh, Gautam and Wang, Yue and Yang, Jiawei and Ivanovic, Boris and Ahn, Sungjin and Pavone, Marco and Che, Tong , booktitle =. Parallelized Spatiotemporal Slot Binding for Videos , year =

  141. [154]

    Elsayed and Aravindh Mahendran and Sjoerd van Steenkiste and Klaus Greff and Michael C

    Gamaleldin F. Elsayed and Aravindh Mahendran and Sjoerd van Steenkiste and Klaus Greff and Michael C. Mozer and Thomas Kipf , booktitle =. SAVi++: Towards End-to-End Object-Centric Learning from Real-World Videos , year =

  142. [155]

    Devlin, Jacob and Chang, Ming-Wei and Lee, Kenton and Toutanova, Kristina , booktitle =

  143. [156]

    Kipf and Elise van der Pol and Max Welling , booktitle =

    Thomas N. Kipf and Elise van der Pol and Max Welling , booktitle =. Contrastive Learning of Structured World Models , year =

  144. [157]

    ArXiv preprint , title =

    Wiedemer, Thadd. ArXiv preprint , title =

  145. [158]

    Hinton , booktitle =

    Sara Sabour and Nicholas Frosst and Geoffrey E. Hinton , booktitle =. Dynamic Routing Between Capsules , year =

  146. [159]

    Hinton and Sara Sabour and Nicholas Frosst , booktitle =

    Geoffrey E. Hinton and Sara Sabour and Nicholas Frosst , booktitle =. Matrix capsules with

  147. [160]

    Capsules with Inverted Dot-Product Attention Routing , year =

    Yao. Capsules with Inverted Dot-Product Attention Routing , year =. 8th International Conference on Learning Representations,

  148. [161]

    Object discovery and representation networks , year =

    H. Object discovery and representation networks , year =. European conference on computer vision , organization =

  149. [162]

    Groupvit: Semantic segmentation emerges from text supervision , year =

    Xu, Jiarui and De Mello, Shalini and Liu, Sifei and Byeon, Wonmin and Breuel, Thomas and Kautz, Jan and Wang, Xiaolong , booktitle =. Groupvit: Semantic segmentation emerges from text supervision , year =

  150. [163]

    Kingma and Max Welling , booktitle =

    Diederik P. Kingma and Max Welling , booktitle =. Auto-Encoding Variational Bayes , year =

  151. [164]

    Siddharth and Yee Whye Teh , booktitle =

    Emile Mathieu and Tom Rainforth and N. Siddharth and Yee Whye Teh , booktitle =. Disentangling Disentanglement in Variational Autoencoders , volume =

  152. [165]

    Cian Eastwood and Christopher K. I. Williams , booktitle =. A Framework for the Quantitative Evaluation of Disentangled Representations , year =

  153. [166]

    Disentangling by Factorising , volume =

    Hyunjik Kim and Andriy Mnih , booktitle =. Disentangling by Factorising , volume =

  154. [167]

    beta-VAE: Learning Basic Visual Concepts with a Constrained Variational Framework , year =

    Irina Higgins and Lo. beta-VAE: Learning Basic Visual Concepts with a Constrained Variational Framework , year =. 5th International Conference on Learning Representations,

  155. [168]

    8th International Conference on Learning Representations,

    Zhixuan Lin and Yi. 8th International Conference on Learning Representations,

  156. [169]

    Investigating object compositionality in generative adversarial networks , volume =

    Van Steenkiste, Sjoerd and Kurach, Karol and Schmidhuber, J. Investigating object compositionality in generative adversarial networks , volume =. Neural Networks , pages =

  157. [170]

    Kosiorek and Oiwi Parker Jones and Ingmar Posner , booktitle =

    Martin Engelcke and Adam R. Kosiorek and Oiwi Parker Jones and Ingmar Posner , booktitle =

  158. [171]

    Masked Autoencoders Are Scalable Vision Learners , year =

    Kaiming He and Xinlei Chen and Saining Xie and Yanghao Li and Piotr Doll. Masked Autoencoders Are Scalable Vision Learners , year =

  159. [172]

    Emerging Properties in Self-Supervised Vision Transformers , year =

    Mathilde Caron and Hugo Touvron and Ishan Misra and Herv. Emerging Properties in Self-Supervised Vision Transformers , year =. 2021

    2021

  160. [173]

    SAT3D: Slot attention transformer for 3D point cloud semantic segmentation , volume =

    Ibrahim, Muhammad and Akhtar, Naveed and Anwar, Saeed and Mian, Ajmal , journal =. SAT3D: Slot attention transformer for 3D point cloud semantic segmentation , volume =

  161. [174]

    A simple and powerful global optimization for unsupervised video object segmentation , year =

    Ponimatkin, Georgy and Samet, Nermin and Xiao, Yang and Du, Yuming and Marlet, Renaud and Lepetit, Vincent , booktitle =. A simple and powerful global optimization for unsupervised video object segmentation , year =

  162. [175]

    Yu and Stephen Lin , booktitle =

    Runtao Liu and Zhirong Wu and Stella X. Yu and Stephen Lin , booktitle =. The Emergence of Objectness: Learning Zero-shot Segmentation from Videos , year =

  163. [176]

    Unsupervised Multi-Object Segmentation by Predicting Probable Motion Patterns , year =

    Laurynas Karazija and Subhabrata Choudhury and Iro Laina and Christian Rupprecht and Andrea Vedaldi , booktitle =. Unsupervised Multi-Object Segmentation by Predicting Probable Motion Patterns , year =

  164. [177]

    Guess What Moves: Unsupervised Video and Image Segmentation by Anticipating Motion , year =

    Subhabrata Choudhury and Laurynas Karazija and Iro Laina and Andrea Vedaldi and Christian Rupprecht , booktitle =. Guess What Moves: Unsupervised Video and Image Segmentation by Anticipating Motion , year =

  165. [178]

    Unsupervised learning of foreground object segmentation , volume =

    Croitoru, Ioana and Bogolin, Simion-Vlad and Leordeanu, Marius , journal =. Unsupervised learning of foreground object segmentation , volume =

  166. [179]

    From ViT Features to Training-free Video Object Segmentation via Streaming-data Mixture Models , year =

    Roy Uziel and Or Dinari and Oren Freifeld , booktitle =. From ViT Features to Training-free Video Object Segmentation via Streaming-data Mixture Models , year =

  167. [180]

    Schwing and Joon

    Ho Kei Cheng and Seoung Wug Oh and Brian Price and Alexander G. Schwing and Joon. Tracking Anything with Decoupled Video Segmentation , year =

  168. [181]

    Fun-sis: A fully unsupervised approach for surgical instrument segmentation , volume =

    Sestini, Luca and Rosa, Benoit and De Momi, Elena and Ferrigno, Giancarlo and Padoy, Nicolas , journal =. Fun-sis: A fully unsupervised approach for surgical instrument segmentation , volume =

  169. [182]

    Gomez and Lukasz Kaiser and Illia Polosukhin , booktitle =

    Ashish Vaswani and Noam Shazeer and Niki Parmar and Jakob Uszkoreit and Llion Jones and Aidan N. Gomez and Lukasz Kaiser and Illia Polosukhin , booktitle =. Attention is All you Need , year =

  170. [183]

    Radford, Alec , title =

  171. [184]

    An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale , year =

    Alexey Dosovitskiy and Lucas Beyer and Alexander Kolesnikov and Dirk Weissenborn and Xiaohua Zhai and Thomas Unterthiner and Mostafa Dehghani and Matthias Minderer and Georg Heigold and Sylvain Gelly and Jakob Uszkoreit and Neil Houlsby , booktitle =. An Image is Worth 16x16 W...

  172. [185]

    Anurag Arnab and Mostafa Dehghani and Georg Heigold and Chen Sun and Mario Lucic and Cordelia Schmid , booktitle =. ViViT:

  173. [186]

    Stacked denoising autoencoders: Learning useful representations in a deep network with a local denoising criterion , year =

    Vincent, Pascal and Larochelle, Hugo and Lajoie, Isabelle and Bengio, Yoshua and Manzagol, Pierre-Antoine and Bottou, Leon , journal =. Stacked denoising autoencoders: Learning useful representations in a deep network with a local denoising criterion , year =

  174. [187]

    Generative Pretraining From Pixels , volume =

    Mark Chen and Alec Radford and Rewon Child and Jeffrey Wu and Heewoo Jun and David Luan and Ilya Sutskever , booktitle =. Generative Pretraining From Pixels , volume =

  175. [188]

    Hangbo Bao and Li Dong and Songhao Piao and Furu Wei , booktitle =. BEiT:

  176. [189]

    PeCo: Perceptual Codebook for

    Xiaoyi Dong and Jianmin Bao and Ting Zhang and Dongdong Chen and Weiming Zhang and Lu Yuan and Dong Chen and Fang Wen and Nenghai Yu and Baining Guo , booktitle =. PeCo: Perceptual Codebook for

  177. [190]

    Masked Feature Prediction for Self-Supervised Visual Pre-Training , year =

    Chen Wei and Haoqi Fan and Saining Xie and Chao. Masked Feature Prediction for Self-Supervised Visual Pre-Training , year =

  178. [191]

    BEVT: BERT pretraining of video transformers , year =

    Wang, Rui and Chen, Dongdong and Wu, Zuxuan and Chen, Yinpeng and Dai, Xiyang and Liu, Mengchen and Jiang, Yu-Gang and Zhou, Luowei and Yuan, Lu , booktitle =. BEVT: BERT pretraining of video transformers , year =

  179. [192]

    Zero-Shot Text-to-Image Generation , volume =

    Aditya Ramesh and Mikhail Pavlov and Gabriel Goh and Scott Gray and Chelsea Voss and Alec Radford and Mark Chen and Ilya Sutskever , booktitle =. Zero-Shot Text-to-Image Generation , volume =

  180. [193]

    VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training , year =

    Zhan Tong and Yibing Song and Jue Wang and Limin Wang , booktitle =. VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training , year =

  181. [194]

    FickleNet: Weakly and Semi-Supervised Semantic Image Segmentation Using Stochastic Inference , year =

    Jungbeom Lee and Eunji Kim and Sungmin Lee and Jangho Lee and Sungroh Yoon , booktitle =. FickleNet: Weakly and Semi-Supervised Semantic Image Segmentation Using Stochastic Inference , year =

  182. [195]

    Self-Erasing Network for Integral Object Attention , year =

    Qibin Hou and Peng. Self-Erasing Network for Integral Object Attention , year =. Advances in Neural Information Processing Systems 31: Annual Conference on Neural Information Processing Systems 2018, NeurIPS 2018, December 3-8, 2018, Montr

    2018

  183. [196]

    Learning to segment moving objects in videos , year =

    Katerina Fragkiadaki and Pablo Arbelaez and Panna Felsen and Jitendra Malik , booktitle =. Learning to segment moving objects in videos , year =

  184. [197]

    Learning Motion Patterns in Videos , year =

    Pavel Tokmakov and Karteek Alahari and Cordelia Schmid , booktitle =. Learning Motion Patterns in Videos , year =

  185. [198]

    Motion-Attentive Transition for Zero-Shot Video Object Segmentation , year =

    Tianfei Zhou and Shunzhou Wang and Yi Zhou and Yazhou Yao and Jianwu Li and Ling Shao , booktitle =. Motion-Attentive Transition for Zero-Shot Video Object Segmentation , year =

  186. [199]

    Spatial broadcast decoder: A simple architecture for learning disentangled representations in vaes , volume =

    Watters, Nicholas and Matthey, Loic and Burgess, Christopher P and Lerchner, Alexander , journal =. Spatial broadcast decoder: A simple architecture for learning disentangled representations in vaes , volume =

  187. [200]

    Mehdi S. M. Sajjadi and Daniel Duckworth and Aravindh Mahendran and Sjoerd van Steenkiste and Filip Pavetic and Mario Lucic and Leonidas J. Guibas and Klaus Greff and Thomas Kipf , booktitle =. Object Scene Representation Transformer , year =

  188. [201]

    Surgical tool classification and localization: results and methods from the MICCAI 2022 SurgToolLoc challenge , volume =

    Zia, Aneeq and Bhattacharyya, Kiran and Liu, Xi and Berniker, Max and Wang, Ziheng and Nespolo, Rogerio and Kondo, Satoshi and Kasai, Satoshi and Hirasawa, Kousuke and Liu, Bo and others , journal =. Surgical tool classification and localization: results and methods from the M...

    2022

  189. [202]

    Endonet: a deep architecture for recognition tasks on laparoscopic videos , volume =

    Twinanda, Andru P and Shehata, Sherif and Mutter, Didier and Marescaux, Jacques and De Mathelin, Michel and Padoy, Nicolas , journal =. Endonet: a deep architecture for recognition tasks on laparoscopic videos , volume =

  190. [203]

    Cholecseg8k: a semantic segmentation dataset for laparoscopic cholecystectomy based on cholec80 , volume =

    Hong, W-Y and Kao, C-L and Kuo, Y-H and Wang, J-R and Chang, W-L and Shih, C-S , journal =. Cholecseg8k: a semantic segmentation dataset for laparoscopic cholecystectomy based on cholec80 , volume =

  191. [204]

    Disentangling spatio-temporal knowledge for weakly supervised object detection and segmentation in surgical video , volume =

    Liao, Guiqiu and Jogan, Matjaz and Koushik, Sai and Eaton, Eric and Hashimoto, Daniel A , journal =. Disentangling spatio-temporal knowledge for weakly supervised object detection and segmentation in surgical video , volume =

  192. [205]

    Multiscale combinatorial grouping for image segmentation and object proposal generation , volume =

    Pont-Tuset, Jordi and Arbelaez, Pablo and Barron, Jonathan T and Marques, Ferran and Malik, Jitendra , journal =. Multiscale combinatorial grouping for image segmentation and object proposal generation , volume =

  193. [206]

    High-resolution image synthesis with latent diffusion models , year =

    Rombach, Robin and Blattmann, Andreas and Lorenz, Dominik and Esser, Patrick and Ommer, Bj. High-resolution image synthesis with latent diffusion models , year =. Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages =

  194. [207]

    Visualizing data using t-SNE

    Van der Maaten, Laurens and Hinton, Geoffrey , journal =. Visualizing data using t-SNE. , volume =

  195. [208]

    Identifiable Object-Centric Representation Learning via Probabilistic Slot Attention , year =

    Kori, Avinash and Locatello, Francesco and Santhirasekaram, Ainkaran and Toni, Francesca and Glocker, Ben and Ribeiro, Fabio De Sousa , journal =. Identifiable Object-Centric Representation Learning via Probabilistic Slot Attention , year =

  196. [209]

    Slot-VLM: Object-Event Slots for Video-Language Modeling , year =

    Xu, Jiaqi and Lan, Cuiling and Xie, Wenxuan and Chen, Xuejin and Lu, Yan , booktitle =. Slot-VLM: Object-Event Slots for Video-Language Modeling , year =

  197. [210]

    Hinton , booktitle =

    Ting Chen and Simon Kornblith and Mohammad Norouzi and Geoffrey E. Hinton , booktitle =. A Simple Framework for Contrastive Learning of Visual Representations , volume =

  198. [211]

    Weakly Supervised Deep Detection Networks , year =

    Hakan Bilen and Andrea Vedaldi , booktitle =. Weakly Supervised Deep Detection Networks , year =

  199. [212]

    An Empirical Study of Training Self-Supervised Vision Transformers , year =

    Xinlei Chen and Saining Xie and Kaiming He , booktitle =. An Empirical Study of Training Self-Supervised Vision Transformers , year =

  200. [213]

    Masked siamese networks for label-efficient learning , year =

    Assran, Mahmoud and Caron, Mathilde and Misra, Ishan and Bojanowski, Piotr and Bordes, Florian and Vincent, Pascal and Joulin, Armand and Rabbat, Mike and Ballas, Nicolas , booktitle =. Masked siamese networks for label-efficient learning , year =

  201. [214]

    A Self-Paced Multiple-Instance Learning Framework for Co-Saliency Detection , year =

    Dingwen Zhang and Deyu Meng and Chao Li and Lu Jiang and Qian Zhao and Junwei Han , booktitle =. A Self-Paced Multiple-Instance Learning Framework for Co-Saliency Detection , year =

  202. [215]

    Learning to Detect Salient Objects with Image-Level Supervision , year =

    Lijun Wang and Huchuan Lu and Yifan Wang and Mengyang Feng and Dong Wang and Baocai Yin and Xiang Ruan , booktitle =. Learning to Detect Salient Objects with Image-Level Supervision , year =

  203. [216]

    Rethinking Transformer-based Set Prediction for Object Detection , year =

    Zhiqing Sun and Shengcao Cao and Yiming Yang and Kris Kitani , booktitle =. Rethinking Transformer-based Set Prediction for Object Detection , year =

  204. [217]

    Complex-Valued Autoencoders for Object Discovery , year =

    L. Complex-Valued Autoencoders for Object Discovery , year =. Transactions on Machine Learning Research , number =

  205. [218]

    Structured Object-Aware Physics Prediction for Video Modeling and Planning , year =

    Jannik Kossen and Karl Stelzner and Marcel Hussing and Claas Voelcker and Kristian Kersting , booktitle =. Structured Object-Aware Physics Prediction for Video Modeling and Planning , year =

  206. [219]

    Learning Transferable Visual Models From Natural Language Supervision , volume =

    Alec Radford and Jong Wook Kim and Chris Hallacy and Aditya Ramesh and Gabriel Goh and Sandhini Agarwal and Girish Sastry and Amanda Askell and Pamela Mishkin and Jack Clark and Gretchen Krueger and Ilya Sutskever , booktitle =. Learning Transferable Visual Models From Natural...

  207. [220]

    OpenAI blog , volume=

    Language models are unsupervised multitask learners , author=. OpenAI blog , volume=

  208. [221]

    Frontiers of Information Technology & Electronic Engineering , volume=

    Federated unsupervised representation learning , author=. Frontiers of Information Technology & Electronic Engineering , volume=. 2023 , publisher=

    2023

  209. [222]

    2021 , eprint=

    Prototype Guided Federated Learning of Visual Feature Representations , author=. 2021 , eprint=

    2021

  210. [223]

    2019 International joint conference on neural networks (IJCNN) , pages=

    Learning private neural language modeling with attentive aggregation , author=. 2019 International joint conference on neural networks (IJCNN) , pages=. 2019 , organization=

    2019

  211. [224]

    Proceedings of the 39th International Conference on Machine Learning , pages =

    Orchestra: Unsupervised Federated Learning via Globally Consistent Clustering , author =. Proceedings of the 39th International Conference on Machine Learning , pages =. 2022 , editor =

    2022

  212. [225]

    Federated Representation Learning Through Clustering , year=

    Miao, Runxuan and Koyuncu, Erdem , booktitle=. Federated Representation Learning Through Clustering , year=

  213. [226]

    European Conference on Computer Vision , pages=

    Fedx: Unsupervised federated learning with cross knowledge distillation , author=. European Conference on Computer Vision , pages=. 2022 , organization=

    2022

  214. [227]

    Advances in neural information processing systems , volume=

    Federated learning from pre-trained models: A contrastive learning approach , author=. Advances in neural information processing systems , volume=

  215. [228]

    Advances in Neural Information Processing Systems , volume=

    Unsupervised object-level representation learning from scene images , author=. Advances in Neural Information Processing Systems , volume=

  216. [229]

    Proceedings of the 2024 International Conference on Intelligent Perception and Pattern Recognition , pages =

    Xiao, He and Dong, Siwen and Xiao, Ningyi , title =. Proceedings of the 2024 International Conference on Intelligent Perception and Pattern Recognition , pages =. 2024 , isbn =. doi:10.1145/3700035.3700050 , abstract =

    2024

  217. [230]

    The Thirteenth International Conference on Learning Representations , year=

    On the transfer of object-centric representation learning , author=. The Thirteenth International Conference on Learning Representations , year=

  218. [231]

    Microsoft

    Lin, Tsung-Yi and Maire, Michael and Belongie, Serge and Hays, James and Perona, Pietro and Ramanan, Deva and Doll. Microsoft. ECCV , pages=

  219. [232]

    International Journal of Computer Vision , volume=

    The Pascal Visual Object Classes Challenge: A Retrospective , author=. International Journal of Computer Vision , volume=

  220. [233]

    ICCV , pages=

    Video Instance Segmentation , author=. ICCV , pages=

  221. [234]

    CVPR , pages=

    Learning Object Class Detectors from Weakly Annotated Video , author=. CVPR , pages=

  222. [235]

    2023 , note=

    Proprietary Thoracic Surgery Dataset , author=. 2023 , note=

    2023

  223. [236]

    Advances in Neural Information Processing Systems , volume=

    Rotating features for object discovery , author=. Advances in Neural Information Processing Systems , volume=

  224. [237]

    arXiv preprint arXiv:2408.09162 , year=

    Zero-shot object-centric representation learning , author=. arXiv preprint arXiv:2408.09162 , year=

    arXiv

  225. [238]

    International Journal of Computer Vision , volume=

    Clip-adapter: Better vision-language models with feature adapters , author=. International Journal of Computer Vision , volume=. 2024 , publisher=

    2024

  226. [239]

    arXiv preprint arXiv:2111.03930 , year=

    Tip-adapter: Training-free clip-adapter for better vision-language modeling , author=. arXiv preprint arXiv:2111.03930 , year=

    arXiv

  227. [240]

    Advances in Neural Information Processing Systems , volume=

    Tinytl: Reduce memory, not parameters for efficient on-device learning , author=. Advances in Neural Information Processing Systems , volume=

  228. [241]

    arXiv preprint arXiv:2106.10199 , year=

    Bitfit: Simple parameter-efficient fine-tuning for transformer-based masked language-models , author=. arXiv preprint arXiv:2106.10199 , year=

  229. [242]

    , author=

    Lora: Low-rank adaptation of large language models. , author=. ICLR , volume=

  230. [243]

    DAGM German Conference on Pattern Recognition , pages=

    Self-masking networks for unsupervised adaptation , author=. DAGM German Conference on Pattern Recognition , pages=. 2024 , organization=

    2024

  231. [244]

    Advances in neural information processing systems , volume=

    Domain separation networks , author=. Advances in neural information processing systems , volume=

  232. [245]

    Advances in neural information processing systems , volume=

    A unified feature disentangler for multi-domain image translation and manipulation , author=. Advances in neural information processing systems , volume=

  233. [246]

    Advances in Neural Information Processing Systems , volume=

    Variational interaction information maximization for cross-domain disentanglement , author=. Advances in Neural Information Processing Systems , volume=

  234. [247]

    Tenth International Conference on Learning Representations 2022 , year=

    Visual representation learning over latent domains , author=. Tenth International Conference on Learning Representations 2022 , year=

    2022

  235. [248]

    2026 , doi=

    Wang, Guankun and Bai, Long and Wang, Junyi and Yuan, Kun and Li, Zhen and Jiang, Tianxu and He, Xiting and Wu, Jinlin and Chen, Zhen and Lei, Zhen and Liu, Hongbin and Wang, Jiazheng and Zhang, Fan and Padoy, Nicolas and Navab, Nassir and Ren, Hongliang , journal=. 2026 , doi=

    2026

  236. [249]

    Advances in Neural Information Processing Systems , volume=

    Visual Instruction Tuning , author=. Advances in Neural Information Processing Systems , volume=

  237. [250]

    Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

    Honeybee: Locality-Enhanced Projector for Multimodal LLM , author=. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

  238. [251]

    Shang, Yuzhang and Cai, Mu and Xu, Bingxin and Lee, Yong Jae and Yan, Yan , booktitle=

  239. [252]

    arXiv preprint arXiv:2405.20985 , year=

    DeCo: Decoupling Token Compression from Semantic Abstraction in Multimodal Large Language Models , author=. arXiv preprint arXiv:2405.20985 , year=

    arXiv

  240. [253]

    Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

    EvoComp: Learning Visual Token Compression for Multimodal Large Language Models via Semantic-Guided Evolutionary Labeling , author=. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

  241. [254]

    Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

    Variation-Aware Vision Token Dropping for Faster Large Vision-Language Models , author=. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

  242. [255]

    arXiv preprint arXiv:2505.17726 , year=

    Slot-MLLM: Object-Centric Visual Tokenization for Multimodal LLM , author=. arXiv preprint arXiv:2505.17726 , year=

    arXiv

  243. [256]

    2024 , doi=

    Ma, Chuofan and Jiang, Yi and Wu, Jiannan and Yuan, Zehuan and Qi, Xiaojuan , booktitle=. 2024 , doi=

    2024

  244. [257]

    arXiv preprint arXiv:2306.14824 , year=

    Kosmos-2: Grounding Multimodal Large Language Models to the World , author=. arXiv preprint arXiv:2306.14824 , year=

    arXiv

  245. [258]

    arXiv preprint arXiv:2306.15195 , year=

    Shikra: Unleashing Multimodal LLM's Referential Dialogue Magic , author=. arXiv preprint arXiv:2306.15195 , year=

    arXiv

  246. [259]

    You, Haoxuan and Zhang, Haotian and Gan, Zhe and Du, Xianzhi and Zhang, Bowen and Wang, Zirui and Cao, Liangliang and Chang, Shih-Fu and Yang, Yinfei , booktitle=

  247. [260]

    Lai, Xin and Tian, Zhuotao and Chen, Yukang and Li, Yanwei and Yuan, Yuhui and Liu, Shu and Jia, Jiaya , booktitle=

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.