Pith. sign in

REVIEW 4 major objections 6 minor 1 cited by

Unlocking Constraints: Source-Free Occlusion-Aware Seamless Segmentation

T0 review · 4 major / 6 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 Source-free panoramic segmentation now matches full-source methods

desk verdict First source-free amodal panoptic segmentation paper with a useful new benchmark, but the headline numbers are compromised by test-set hyperparameter selection. read the letter →

arxiv 2506.21198 v2 pith:LDU4TPRZ submitted 2025-06-26 cs.CV cs.ROeess.IV

Yihong Cao , Jiaming Zhang , Xu Zheng , Hao Shi , Kunyu Peng , Hang Liu , Kailun Yang , Hui Zhang This is my paper · ORCID
classification cs.CVcs.ROeess.IV
keywords source-freedomainadaptationpanoramicsegmentationamodalinstancepanopticpseudo-labelingclass-wiseself-tuningthresholdssynthetic-to-realocclusion-awarereasoning
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

Panoramic scene understanding normally needs either labeled panoramic data or access to the labeled pinhole data used to train the source model. This paper introduces a stricter task, source-free occlusion-aware seamless segmentation (SFOASS), in which neither the source images nor the target labels are available during adaptation, and proposes the first solution, UNLOCK. The paper claims that a pinhole-trained model can be adapted to unlabeled 360-degree imagery purely through pseudo-labeling and a mixing strategy, and backs this with state-of-the-art scores of 10.9 in mAAP and 11.6 in mAP, plus a +4.3 gain in mAPQ over the source-only baseline on the real-to-real benchmark. The contribution matters because privacy, storage, and commercial restrictions often make the source dataset inaccessible, and the results suggest that giving up source access need not cost performance.

What carries the argument

Two interacting modules carry the argument. Omni Pseudo-Labeling Learning (OPLL) combines class-wise self-tuning thresholds with semantic-to-instance mask revision: the semantic branch's Thing mask is intersected with each instance-level prediction, so objects the local instance branch mislabels can be corrected, and an uncertainty-guided binary cross-entropy loss masks out low-confidence regions so wrong pseudo-labels are not penalized as false background. Amodal-Driven Contextual Learning (ADCL) builds a pool of high-confidence amodal object samples and mixes them into training images, zeroing only the overlapping (potentially occluded) regions while keeping the full amodal mask as the target label, which preserves the complete shape of objects without contaminating scene context.

What would settle it

Take a target panorama set with deliberately degraded or scrambled semantic predictions (e.g., corrupt a fraction of the semantic Thing masks) and rerun UNLOCK while keeping the instance branches untouched: if mAPQ and mAAP drop to or below the source-only baseline as corruption increases, the semantic-revision premise is doing the claimed work. Alternatively, compare UNLOCK against a variant that revises instance masks using a different source of object evidence, such as instance branch self-consistency, on a heavy-distortion target; if that variant matches or beats UNLOCK, the semantic-trust assumption is not necessary.

Watch

Extended reading notes

Core claim

The central claim is that source-free adaptation is sufficient for occlusion-aware seamless segmentation on panoramic images. UNLOCK generates omni pseudo-labels from all three branches of the source model, using class-wise self-tuning thresholds that adapt per category rather than one global cutoff, and refines instance-level masks with the semantic branch's Thing mask. It then builds an amodal-driven object pool of the most confident amodal instance predictions and uses a spatial-aware mixing strategy that pastes full object shapes while zeroing only their overlapping regions, so the model sees complete occluded objects without corrupting contextual cues. On KITTI360-APS->BlendPASS the method reaches 26.4 mAPQ, 10.9 mAAP, and 11.6 mAP, surpassing UDA methods that use all 12K source image-label pairs, and on the new AmodalSynthDrive->BlendPASS synthetic-to-real benchmark it also leads across the five segmentation metrics.

Load-bearing premise

The load-bearing premise is that on the target panoramic domain the source model's semantic branch is more trustworthy than its instance-level branches, so instance pseudo-labels can safely be revised by intersecting them with the semantic Thing mask; if a semantic prediction is wrong on heavily distorted or unfamiliar classes, correct instance detections can be discarded or corrupted.

Editorial extensions

If this is right

  • Source-free adaptation can serve as a drop-in replacement when source data is restricted, reducing storage and privacy burdens.
  • Synthetic-to-real transfer works for occlusion-aware segmentation, so simulators can be used to train panoramic models without real pinhole labels.
  • The method generalizes beyond OASS: applied to semantic-only panoramic adaptation (C2D benchmark) it improves mIoU to 54.55, indicating the pseudo-labeling strategy transfers.
  • The class-wise thresholding component addresses the imbalance in source-model confidence across categories, a common failure mode of fixed-threshold self-training.
  • UNLOCK is backbone-agnostic, showing consistent gains across different architectures, so the adaptation strategy can be composed with stronger base models.

Reading between the lines

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

  • Beyond the paper: the principle of zeroing only overlapping regions while keeping full amodal masks suggests a general recipe for amodal data augmentation, and applying it to other amodal datasets such as indoor scenes is a direct test.
  • Beyond the paper: if source-free adaptation matches UDA here, the value of source data may mostly lie in pre-training, and future work could investigate how much source data is actually needed before it stops helping.
  • Beyond the paper: the reported failure case of vehicles behind sparse fences suggests that structured occluders that break spatial continuity, not just dense ones, are the remaining hard mode; a stress-test suite built from such occluders would be a useful benchmark.
  • Beyond the paper: the SFOASS task formulation could become a template for other source-free multi-branch dense prediction settings, such as 3D detection with instance and amodal branches.
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

4 major / 6 minor

Summary. The manuscript introduces Source-Free Occlusion-Aware Seamless Segmentation (SFOASS), a task setting in which a model trained on labeled pinhole images must adapt to unlabeled panoramic images without access to the source data. The proposed UNLOCK framework combines Omni Pseudo-Labeling Learning (OPLL), which generates class-wise thresholded pseudo-labels for semantic, instance, and amodal instance branches, with Amodal-Driven Context Learning (ADCL), which builds an object pool from high-confidence amodal predictions and pastes objects into training images with zeroed occlusion regions. Experiments are reported on KITTI360-APS→BlendPASS (K2B), a newly introduced AmodalSynthDrive→BlendPASS (A2B) benchmark, and Cityscapes-to-DensePASS (C2D) for semantic segmentation. The paper claims that UNLOCK achieves state-of-the-art mAAP of 10.9 and mAP of 11.6, and an absolute improvement of +4.3 mAPQ over the source-only model, reaching performance comparable to source-dependent UDA methods.

Significance. If the reported results are reliable, the paper would make a useful contribution by extending source-free domain adaptation to the relatively new occlusion-aware seamless segmentation task and by providing a synthetic-to-real benchmark for that task. The task formulation is clear, the method is described in enough detail to be implemented, and the authors commit to releasing data and code. The paper also includes ablations of the two main components and additional results on panoramic semantic segmentation. However, the central quantitative claims are currently undermined by the apparent use of the target test set for hyperparameter selection and by the absence of any uncertainty quantification. The claimed state-of-the-art numbers and the headline +4.3 mAPQ gain therefore cannot yet be taken as evidence for the method's intrinsic source-free adaptation ability; they must be re-established under a clean validation protocol.

major comments (4)
  1. [Section 9, Tables 18-19] The hyperparameter sweeps in Section 9 are performed directly on the K2B benchmark, whose only labeled target data are the 100 test images (Sec. 4.1), and the reported final configurations are selected as the gray rows that maximize mAPQ. Table 18 shows mAPQ ranging from 24.80 to 25.84 as tau'_fix and tau'_per vary, and Table 19 shows mAPQ ranging from 24.83 to 25.84 as R varies, with the chosen R=10 at the peak. This is effectively test-label hyperparameter selection, and it means the reported +4.3 mAPQ improvement and the mAAP/mAP state-of-the-art numbers partly reflect tuning on the test set rather than the method's intrinsic performance. Please introduce a proper validation split (or cross-validation) for threshold selection, or justify fixed thresholds by a source-only or model-free criterion, and report final results on the held-out test set with the chosen configuration.
  2. [Section 4.3, Tables 1-7] All experimental results are reported as single runs without error bars or multiple seeds. Since the BlendPASS target test set contains only 100 images, the reported margins on instance-level metrics (e.g., mAAP 10.9 vs. 10.5 for the source-only baseline, or mAP 11.6 vs. 11.3 for the EDAPS baseline) are small and could easily lie within run-to-run variation. The paper should report means and standard deviations over at least three seeds, and ideally a significance test, for the main tables and the component ablations, before claims such as 'state-of-the-art' or '+4.3 improvement' are substantiated.
  3. [Section 3.2, Eqs. (2)-(3)] OPLL revises every instance and amodal instance mask by intersecting it with the semantic branch's Thing mask. This assumes the semantic branch is more reliable than the instance branches on the target panoramic domain, but the paper provides no evidence for this assumption on classes where the semantic branch is weak. For example, in Table 1 the semantic branch achieves mIoU of only 18.3 for cyclists and 0.0 for two-wheelers on the K2B benchmark, so the semantic Thing mask may suppress correct instance predictions for exactly the classes where pseudo-labels are most needed. Please quantify how many instance proposals are discarded by Eq. (3), report per-class agreement between the semantic and instance branches, and add an ablation that removes or softens this semantic masking step.
  4. [Section 4.3 and Section 8] The comparison to UDA baselines is not yet framed fairly. UNLOCK's thresholds are tuned on the K2B target test set in Section 9, while the reproduced UDA baselines on A2B are stated in Section 8 to use hyperparameters from their original papers. It is not stated whether the same K2B-tuned thresholds are transferred unchanged to the A2B benchmark or whether the baselines were given any comparable validation-based tuning. This asymmetry could inflate the reported advantage over source-dependent methods. Please state the exact hyperparameters used for UNLOCK on A2B, and if they are the K2B-tuned values, either justify the transfer or re-tune all methods on a common validation protocol before computing the final numbers.
minor comments (6)
  1. [Section 4.2] The text says 'The analysis of these thresholds is provided in the supplementary material,' but the threshold analysis appears in Section 9 of the main paper; this cross-reference should be corrected.
  2. [Table 2] The right half of Table 2 labels the Instance Segmentation metric as 'mAAP for Instance Segmentation (IS)'; this should be 'mAP' to match the metric name used in the abstract and Figure 1.
  3. [Section 3.2, Eq. (6)] Equation (6) is notationally unclear: the expressions '1_{P_j ...}' appear to use an undefined summation symbol, and the equality mixes set-level and pixel-level notation. The uncertainty region should be defined with explicit spatial indices so that the intersection and the indicator functions are unambiguous.
  4. [Figure 13] The caption of Figure 13 is identical to the caption of Figure 12 and describes a failure case, but the figure actually shows qualitative examples of the amodal-driven object pool; the caption should be rewritten accordingly.
  5. [Table 1] Several entries in Table 1 contain corrupted formatting with multiple concatenated numbers (e.g., '10.101.3' for DATR and '15.304.241.1' for Trans4PASS under mAPQ); the rendered table should be checked so that each cell contains exactly one number.
  6. [Section 4.3] The phrase 'comparable results' with respect to UnmaskFormer should be qualified, since Table 1 shows UNLOCK is about one point lower on mPQ and mIoU while being higher on instance metrics; the current wording is stronger than the evidence in the table.

Circularity Check

1 steps flagged · score 6.0 of 10

Reported SOTA numbers are selected on the 100-image test set via Section 9 threshold sweeps, making the headline improvements self-confirming.

  1. fitted input called prediction [Appendix Section 9 (Analysis for Hyper-parameters), Tables 9-19; Section 4.1 dataset description]
    "Since these metrics do not always vary consistently, we prioritize mAPQ, which evaluates amodal panoramic segmentation (including both Stuff and amodal-level Thing classes), while also taking into account the overall performance across all metrics. ... only the two thresholds associated with the branch under investigation were varied, while the thresholds for the other branches were held constant at their final adopted values, as indicated by the gray background in the tables."

    The paper describes no validation split for BlendPASS: Section 4.1 states it 'consists of 2,000 unlabeled training panoramic images and 100 labeled test panoramic images.' Section 9 sweeps thresholds and reports mAPQ, mAAP, mAP, etc. on the KITTI360-APS→BlendPASS benchmark, i.e., on those 100 test labels. The gray rows mark the 'final adopted values' selected by prioritizing mAPQ. The reported headline numbers (mAPQ 26.4, mAAP 10.9, mAP 11.6, +4.3 mAPQ over source-only) are therefore the test-optimal configuration, not an independent evaluation of a fixed method. The improvement is statistically forced by test-label-based hyperparameter selection and then presented as a prediction; baselines are not shown to have received equivalent test-label tuning.

full rationale

The method itself contains no equation-level tautology: OPLL generates pseudo-labels from the source model via class-wise thresholds and uncertainty-weighted BCE, and ADCL performs spatial-aware mixing; these are normal self-training and augmentation operations, not definitions that presuppose the reported metrics. The central circularity is in the evaluation chain. Because BlendPASS provides only 2,000 unlabeled training images and 100 labeled test images, and no validation or development split is mentioned, the Appendix's threshold sweeps (Tables 9-19) directly optimize the reported test metrics. The gray-highlighted configurations are declared final based on mAPQ while also considering overall performance, so the paper's SOTA claims are self-confirming: the model is evaluated on the same labels used to pick its hyperparameters. This does not invalidate every component—the backbone comparison and the A2B results still provide some independent signal—but the central Real-to-Real quantitative claim, including the +4.3 mAPQ improvement and the state-of-the-art mAAP/mAP numbers, reduces to a fitted choice rather than an out-of-sample prediction. Under the review's pattern 'fitted input called prediction,' this warrants a score of 6: partial circularity in the central claim, while the core adaptation machinery retains independent content.

Assumptions & free parameters 5 free parameters · 3 assumptions · 0 invented entities

The central claim rests on the reliability of model-generated pseudo-labels and on hyperparameters chosen against the target benchmark. No new physical entities are introduced.

free parameters (5)
  • tau_fix and tau_per for amodal instance pseudo-labels = 0.3 and 0.5
    Chosen by sweeping on K2B (Table 9) to maximize target metrics; no validation protocol stated.
  • tau_fix and tau_per for instance pseudo-labels = 0.5 and 0.3
    Chosen via target-metric sweeps (Table 12).
  • tau_fix and tau_per for semantic pseudo-labels = 0.5 and 0.8
    Chosen via target-metric sweeps (Table 15).
  • tau'_fix and tau'_per for ADCL object pool = 0.95 and 0.1 in the main setup; Table 18 shows 0.90 and 0.15 as best without OPLL
    Hyperparameters for the amodal-driven object pool; selected by sweeping on target metrics (Table 18).
  • R, number of pasted objects per mixed image = 10
    Selected via sweep on target metrics (Table 19).
assumptions (3)
  • domain assumption The source model's predictions on target panoramic images are informative enough to serve as pseudo-labels after thresholding.
    The whole OPLL self-training procedure depends on this; the paper does not provide an independent check of pseudo-label accuracy.
  • domain assumption The semantic branch is more reliable than instance branches on the target domain, so instance masks can be revised using the semantic thing mask (Eq. 2-3).
    This is explicit in Section 3.2 and is load-bearing for OPLL.
  • ad hoc to paper Zeroing only the overlapping regions of pasted objects (ADCL) preserves useful context while teaching amodal completion.
    This is a design choice validated only empirically on one benchmark; it is not derived from a general principle.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Unlocking Constraints: Source-Free Occlusion-Aware Seamless Segmentation." pith.science (2026). https://pith.science/paper/LDU4TPRZ

@misc{pith2026250621198,
  author       = {Pith},
  title        = {Pith review of: Unlocking Constraints: Source-Free Occlusion-Aware Seamless Segmentation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LDU4TPRZ}},
  note         = {Machine review of arXiv:2506.21198}
}
read the original abstract

Panoramic image processing is essential for omni-context perception, yet faces constraints like distortions, perspective occlusions, and limited annotations. Previous unsupervised domain adaptation methods transfer knowledge from labeled pinhole data to unlabeled panoramic images, but they require access to source pinhole data. To address these, we introduce a more practical task, i.e., Source-Free Occlusion-Aware Seamless Segmentation (SFOASS), and propose its first solution, called UNconstrained Learning Omni-Context Knowledge (UNLOCK). Specifically, UNLOCK includes two key modules: Omni Pseudo-Labeling Learning and Amodal-Driven Context Learning. While adapting without relying on source data or target labels, this framework enhances models to achieve segmentation with 360{\deg} viewpoint coverage and occlusion-aware reasoning. Furthermore, we benchmark the proposed SFOASS task through both real-to-real and synthetic-to-real adaptation settings. Experimental results show that our source-free method achieves performance comparable to source-dependent methods, yielding state-of-the-art scores of 10.9 in mAAP and 11.6 in mAP, along with an absolute improvement of +4.3 in mAPQ over the source-only method. All data and code will be made publicly available at https://github.com/yihong-97/UNLOCK.

Figures

Figures reproduced from arXiv: 2506.21198 by the authors.

Figure 1
Figure 1. UNLOCK framework (a) solves the Source-Free Occlusion-Aware Seamless Segmentation (SFOASS), enabling segmentation with 360° viewpoint coverage and occlusion-aware reasoning while adapting without requiring source data and target labels, and (b) outperforms existing SFDA methods on the Real-to-Real scenario [7] across all five segmentation metrics (i.e., mAPQ for amodal panoptic, mPQ for panoptic, mIoU for semantic, … view at source ↗
Figure 2
Figure 2. Illustration of the proposed UNLOCK. To address the challenge of the inaccessible source domain and target labels in SFOASS, [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Solution to the challenge of predictions from the source [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (9 more)
Figure 4
Figure 4. Figure 4: Comparison of object extraction methods: Amodal [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: Example images from the datasets [7, 49, 58] and the class distribution visualized via t-SNE [67]. 4.2. Experiment Setups Following [7], we use the proposed architecture as the backbone of our UNLOCK and the reproduced 360SFUDA++ [96]. In the A2B benchmark, we reproduc…
Figure 6
Figure 6. Figure 6: Visualization results. From top to bottom are (a) Image, (b) GT, (c) Source-only, and (d) 360SFUDA++ [96], and (e) UNLOCK. set to 10. The analysis of these thresholds is provided in the supplementary material. Our experiments are conducted on a single NVIDIA GTX 3090 G…
Figure 8
Figure 8. Figure 8: Workflow of the training and inference phases of the [PITH_FULL_IMAGE:figures/full_fig_p013_8.png]
Figure 7
Figure 7. Figure 7: Comparison of Task Settings. However, as a task based on Unsupervised Domain Adaptation (UDA), OASS requires simultaneous access to the data of both the source domain and the target domain during the adaptation process. This presents challenges in scenarios with data p…
Figure 9
Figure 9. Figure 9: Visualization for Amodal Panoptic Segmentation on KITTI360-APS→BlendPASS benchmark. From top to bottom are (a) Image, (b) GT, (c) DATR [94], (d) Trans4PASS [86], (e) EDAPS [56], (f) UnmaskFormer [7], (g) Source-only, (h) 360SFUDA++ [96], and (i) UNLOCK (Ours). 10.3. Qu…
Figure 10
Figure 10. Figure 10: Visualization for Amodal Panoptic Segmentation on AmodalSynthDrive→BlendPASS benchmark. From top to bottom are (a) Image, (b) GT, (c) DATR [94], (d) Trans4PASS [86], (e) EDAPS [56], (f) UnmaskFormer [7], (g) Source-only, (h) 360SFUDA++ [96], and (i) UNLOCK (Ours). (b)…
Figure 11
Figure 11. Figure 11: Visualization for Semantic (Left) and Panoptic (Right) Segmentation on KITTI360-APS→BlendPASS benchmark. From top to bottom are (a) Image, (b) GT, (c) DATR [94], (d) Trans4PASS [86], (e) UniDAPS [87], (f) EDAPS [56], (g) UnmaskFormer [7], (h) Source-only, (i) 360SFUDA…
Figure 12
Figure 12. Figure 12: Example of a failure case where UNLOCK misses vehicles partially occluded by sparse fences [PITH_FULL_IMAGE:figures/full_fig_p019_12.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Omnidirectional Spatial Modeling from Correlated Panoramas

    cs.CV 2025-09 conditional novelty 5.0 of 10

    The authors create a cross-frame panoramic VQA benchmark from 3D scene data and show that GRPO fine-tuning of Qwen2.5-VL raises its score on that benchmark.

Reference graph

Works this paper leans on

107 extracted references · 78 canonical work pages · cited by 1 Pith paper

  1. [1]

    A survey of represen- tation learning, optimization strategies, and applications for omnidirectional vision

    Hao Ai, Zidong Cao, and Lin Wang. A survey of represen- tation learning, optimization strategies, and applications for omnidirectional vision. International Journal of Computer Vision, 2025. 1

    2025

  2. [2]

    Self-training via Metric Learning for Source-Free Domain Adaptation of Semantic Segmentation

    Ibrahim Batuhan Akkaya and Ugur Halici. Self-training via metric learning for source-free domain adaptation of semantic segmentation. arXiv preprint arXiv:2212.04227,

    work page Pith review arXiv

  3. [3]

    PanNote: an automatic tool for panoramic im- age annotation of people’s positions

    Alberto Bacchin, Leonardo Barcellona, Sepideh Sham- sizadeh, Emilio Olivastri, Alberto Pretto, and Emanuele Menegatti. PanNote: an automatic tool for panoramic im- age annotation of people’s positions. In ICRA, 2024. 1

    2024

  4. [4]

    Un- seen object amodal instance segmentation via hierarchical occlusion modeling

    Seunghyeok Back, Joosoon Lee, Taewon Kim, Sangjun Noh, Raeyoung Kang, Seongho Bak, and Kyoobin Lee. Un- seen object amodal instance segmentation via hierarchical occlusion modeling. In ICRA, 2022. 2

    2022

  5. [5]

    Amodal cityscapes: A new dataset, its generation, and an amodal semantic segmentation challenge baseline

    Jasmin Breitenstein and Tim Fingscheidt. Amodal cityscapes: A new dataset, its generation, and an amodal semantic segmentation challenge baseline. In IV, 2022. 2

    2022

  6. [6]

    Towards source-free domain adaptive semantic segmentation via importance-aware and prototype-contrast learning

    Yihong Cao, Hui Zhang, Xiao Lu, Zheng Xiao, Kailun Yang, and Yaonan Wang. Towards source-free domain adaptive semantic segmentation via importance-aware and prototype-contrast learning. IEEE Transactions on Intelli- gent Vehicles, 2024. 3

    2024

  7. [7]

    Occlusion-aware seamless segmentation

    Yihong Cao, Jiaming Zhang, Hao Shi, Kunyu Peng, Yuhongxuan Zhang, Hui Zhang, Rainer Stiefelhagen, and Kailun Yang. Occlusion-aware seamless segmentation. In ECCV, 2024. 1, 2, 3, 5, 6, 7, 8, 13, 14, 16, 17, 18

    2024

  8. [8]

    Amodal instance segmentation via prior- guided expansion

    Junjie Chen, Li Niu, Jianfu Zhang, Jianlou Si, Chen Qian, and Liqing Zhang. Amodal instance segmentation via prior- guided expansion. In AAAI, 2023. 2

    2023

Show all 107 references
  1. [9]

    Encoder-decoder with atrous separable convolution for semantic image segmenta- tion

    Liang-Chieh Chen, Yukun Zhu, George Papandreou, Flo- rian Schroff, and Hartwig Adam. Encoder-decoder with atrous separable convolution for semantic image segmenta- tion. In ECCV, 2018. 1

    2018

  2. [10]

    The cityscapes dataset for semantic urban scene understanding

    Marius Cordts, Mohamed Omran, Sebastian Ramos, Timo Rehfeld, Markus Enzweiler, Rodrigo Benenson, Uwe Franke, Stefan Roth, and Bernt Schiele. The cityscapes dataset for semantic urban scene understanding. In CVPR,

  3. [11]

    Convolutional fea- ture masking for joint object and stuff segmentation

    Jifeng Dai, Kaiming He, and Jian Sun. Convolutional fea- ture masking for joint object and stuff segmentation. In CVPR, 2015. 2

    2015

  4. [12]

    360° camera alignment via segmentation

    Benjamin Davidson, Mohsan S Alvi, and Jo˜ao F Henriques. 360° camera alignment via segmentation. In ECCV, 2020. 1

    2020

  5. [13]

    CNN based semantic segmenta- tion for urban traffic scenes using fisheye camera

    Liuyuan Deng, Ming Yang, Yeqiang Qian, Chunxiang Wang, and Bing Wang. CNN based semantic segmenta- tion for urban traffic scenes using fisheye camera. In IV,

  6. [14]

    CARLA: An open urban driving simulator

    Alexey Dosovitskiy, German Ros, Felipe Codevilla, Anto- nio Lopez, and Vladlen Koltun. CARLA: An open urban driving simulator. In CoRL, 2017. 6

    2017

  7. [15]

    Source-free domain adaptation for point cloud semantic segmentation

    Jianshe Duan, Yachao Zhang, and Yanyun Qu. Source-free domain adaptation for point cloud semantic segmentation. In ICME, 2024. 3

    2024

  8. [16]

    Rethinking amodal video segmentation from learning supervised sig- nals with object-centric representation

    Ke Fan, Jingshi Lei, Xuelin Qian, Miaopeng Yu, Tianjun Xiao, Tong He, Zheng Zhang, and Yanwei Fu. Rethinking amodal video segmentation from learning supervised sig- nals with object-centric representation. In ICCV, 2023. 2

    2023

  9. [17]

    Learning to see the invisible: End-to-end trainable amodal instance seg- mentation

    Patrick Follmann, Rebecca K ¨onig, Philipp H ¨artinger, Michael Klostermann, and Tobias B ¨ottger. Learning to see the invisible: End-to-end trainable amodal instance seg- mentation. In WACV, 2019. 2

    2019

  10. [18]

    PanopticNeRF-360: Panoramic 3D-to-2D label transfer in urban scenes

    Xiao Fu, Shangzhan Zhang, Tianrun Chen, Yichong Lu, Xiaowei Zhou, Andreas Geiger, and Yiyi Liao. PanopticNeRF-360: Panoramic 3D-to-2D label transfer in urban scenes. arXiv preprint arXiv:2309.10815, 2023. 2

    2023 arXiv

  11. [19]

    Coarse-to-fine amodal segmentation with shape prior

    Jianxiong Gao, Xuelin Qian, Yikai Wang, Tianjun Xiao, Tong He, Zheng Zhang, and Yanwei Fu. Coarse-to-fine amodal segmentation with shape prior. In ICCV, 2023. 2

    2023

  12. [20]

    Review on panoramic imaging and its applications in scene understanding

    Shaohua Gao, Kailun Yang, Hao Shi, Kaiwei Wang, and Jian Bai. Review on panoramic imaging and its applications in scene understanding. IEEE Transactions on Instrumen- tation and Measurement, 2022. 1

    2022

  13. [21]

    Geometric unsupervised domain adapta- tion for semantic segmentation

    Vitor Campanholo Guizilini, Jie Li, Rares Ambrus, and Adrien Gaidon. Geometric unsupervised domain adapta- tion for semantic segmentation. In ICCV, 2021. 3

    2021

  14. [22]

    SegNeXt: Rethinking convolutional attention design for semantic segmentation

    Meng-Hao Guo, Cheng-Ze Lu, Qibin Hou, Zhengning Liu, Ming-Ming Cheng, and Shi-Min Hu. SegNeXt: Rethinking convolutional attention design for semantic segmentation. In NeurIPS, 2022. 1

    2022

  15. [23]

    Single frame se- mantic segmentation using multi-modal spherical images

    Suresh Guttikonda and Jason Rambach. Single frame se- mantic segmentation using multi-modal spherical images. In WACV, 2024. 2

    2024

  16. [24]

    Simultaneous detection and segmentation

    Bharath Hariharan, Pablo Arbel ´aez, Ross Girshick, and Ji- tendra Malik. Simultaneous detection and segmentation. In ECCV, 2014. 2

    2014

  17. [25]

    DAFormer: Improving network architectures and training strategies for domain-adaptive semantic segmentation

    Lukas Hoyer, Dengxin Dai, and Luc Van Gool. DAFormer: Improving network architectures and training strategies for domain-adaptive semantic segmentation. In CVPR, 2022. 2

    2022

  18. [26]

    HRDA: Context-aware high-resolution domain-adaptive semantic segmentation

    Lukas Hoyer, Dengxin Dai, and Luc Van Gool. HRDA: Context-aware high-resolution domain-adaptive semantic segmentation. In ECCV, 2022. 2

    2022

  19. [27]

    Yuan-Ting Hu, Hong-Shuo Chen, Kexin Hui, Jia-Bin Huang, and Alexander G. Schwing. SAIL-VOS: Semantic amodal instance level video object segmentation - A syn- thetic dataset and baselines. In CVPR, 2019. 2

    2019

  20. [28]

    Cross-view regularization for domain adaptive panoptic segmentation

    Jiaxing Huang, Dayan Guan, Aoran Xiao, and Shijian Lu. Cross-view regularization for domain adaptive panoptic segmentation. In CVPR, 2021. 3

    2021

  21. [29]

    DaDA: Distortion- aware domain adaptation for unsupervised semantic seg- mentation

    Sujin Jang, Joohan Na, and Dokwan Oh. DaDA: Distortion- aware domain adaptation for unsupervised semantic seg- mentation. In NeurIPS, 2022. 2

    2022

  22. [30]

    Panoramic panoptic segmentation: Towards complete sur- rounding understanding via unsupervised contrastive learn- ing

    Alexander Jaus, Kailun Yang, and Rainer Stiefelhagen. Panoramic panoptic segmentation: Towards complete sur- rounding understanding via unsupervised contrastive learn- ing. In IV, 2021. 3

    2021

  23. [31]

    Panoramic panoptic segmentation: Insights into surround- ing parsing for mobile agents via unsupervised contrastive learning

    Alexander Jaus, Kailun Yang, and Rainer Stiefelhagen. Panoramic panoptic segmentation: Insights into surround- ing parsing for mobile agents via unsupervised contrastive learning. IEEE Transactions on Intelligent Transportation Systems, 2023. 2, 3

    2023

  24. [32]

    Multi-source domain adap- tation for panoramic semantic segmentation

    Jing Jiang, Sicheng Zhao, Jiankun Zhu, Wenbo Tang, Zhaopan Xu, Jidong Yang, Guoping Liu, Tengfei Xing, Pengfei Xu, and Hongxun Yao. Multi-source domain adap- tation for panoramic semantic segmentation. Information Fusion, 2025. 3

    2025

  25. [33]

    Domain generalization via balancing training difficulty and model capability

    Xueying Jiang, Jiaxing Huang, Sheng Jin, and Shijian Lu. Domain generalization via balancing training difficulty and model capability. In ICCV, 2023. 2

    2023

  26. [34]

    C-SFDA: A curriculum learning aided self-training framework for efficient source free domain adaptation

    Nazmul Karim, Niluthpol Chowdhury Mithun, Abhinav Rajvanshi, Han-pang Chiu, Supun Samarasekera, and Nazanin Rahnavard. C-SFDA: A curriculum learning aided self-training framework for efficient source free domain adaptation. In CVPR, 2023. 3

    2023

  27. [35]

    Amodal instance segmentation

    Ke Li and Jitendra Malik. Amodal instance segmentation. In ECCV, 2016. 2

    2016

  28. [36]

    SGAT4PASS: Spherical geometry-aware transformer for panoramic semantic segmentation

    Xuewei Li, Tao Wu, Zhongang Qi, Gaoang Wang, Ying Shan, and Xi Li. SGAT4PASS: Spherical geometry-aware transformer for panoramic semantic segmentation. In IJ- CAI, 2023. 2

    2023

  29. [37]

    2D amodal instance segmentation guided by 3D shape prior

    Zhixuan Li, Weining Ye, Tingting Jiang, and Tiejun Huang. 2D amodal instance segmentation guided by 3D shape prior. In ECCV, 2022. 2

    2022

  30. [38]

    GIN: Generative invariant shape prior for amodal instance segmentation

    Zhixuan Li, Weining Ye, Tingting Jiang, and Tiejun Huang. GIN: Generative invariant shape prior for amodal instance segmentation. IEEE Transactions on Multimedia, 2023. 2

    2023

  31. [39]

    MUV A: A new large-scale benchmark for multi-view amodal instance segmentation in the shopping scenario

    Zhixuan Li, Weining Ye, Juan Terven, Zachary Bennett, Ying Zheng, Tingting Jiang, and Tiejun Huang. MUV A: A new large-scale benchmark for multi-view amodal instance segmentation in the shopping scenario. In ICCV, 2023. 2

    2023

  32. [40]

    KITTI-360: A novel dataset and benchmarks for urban scene understand- ing in 2D and 3D

    Yiyi Liao, Jun Xie, and Andreas Geiger. KITTI-360: A novel dataset and benchmarks for urban scene understand- ing in 2D and 3D. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2023. 6

    2023

  33. [41]

    Source-free domain adaptation for semantic segmentation

    Yuang Liu, Wei Zhang, and Jun Wang. Source-free domain adaptation for semantic segmentation. In CVPR, 2021. 2, 3, 8

    2021

  34. [42]

    BLADE: Box-level supervised amodal segmentation through di- rected expansion

    Zhaochen Liu, Zhixuan Li, and Tingting Jiang. BLADE: Box-level supervised amodal segmentation through di- rected expansion. In AAAI, 2024. 2

    2024

  35. [43]

    Decoupled weight decay regularization

    Ilya Loshchilov and Frank Hutter. Decoupled weight decay regularization. In ICLR, 2019. 6

    2019

  36. [44]

    Hospedales

    Zhihe Lu, Da Li, Yi-Zhe Song, Tao Xiang, and Timothy M. Hospedales. Uncertainty-aware source-free domain adap- tive semantic segmentation. IEEE Transactions on Image Processing, 2023. 3

    2023

  37. [45]

    DensePASS: Dense panoramic semantic segmentation via unsupervised domain adaptation with attention-augmented context exchange

    Chaoxiang Ma, Jiaming Zhang, Kailun Yang, Alina Roitberg, and Rainer Stiefelhagen. DensePASS: Dense panoramic semantic segmentation via unsupervised domain adaptation with attention-augmented context exchange. In ITSC, 2021. 1

    2021

  38. [46]

    Language-guided instance-aware domain-adaptive panoptic segmentation

    Elham Amin Mansour, Ozan Unal, Suman Saha, Benjamin Bejar, and Luc Van Gool. Language-guided instance-aware domain-adaptive panoptic segmentation. arXiv preprint arXiv:2404.03799, 2024. 3

    2024 arXiv

  39. [47]

    MC- PanDA: Mask confidence for panoptic domain adaptation

    Ivan Martinovic, Josip Saric, and Sinisa Segvic. MC- PanDA: Mask confidence for panoptic domain adaptation. In ECCV, 2024. 3

    2024

  40. [48]

    Waymo open dataset: Panoramic video panoptic segmenta- tion

    Jieru Mei, Alex Zihao Zhu, Xinchen Yan, Hang Yan, Siyuan Qiao, Liang-Chieh Chen, and Henrik Kretzschmar. Waymo open dataset: Panoramic video panoptic segmenta- tion. In ECCV, 2022. 2

    2022

  41. [49]

    Amodal panoptic seg- mentation

    Rohit Mohan and Abhinav Valada. Amodal panoptic seg- mentation. In CVPR, 2022. 3, 6, 13

    2022

  42. [50]

    Perceiving the invis- ible: Proposal-free amodal panoptic segmentation

    Rohit Mohan and Abhinav Valada. Perceiving the invis- ible: Proposal-free amodal panoptic segmentation. IEEE Robotics and Automation Letters, 2022. 3

    2022

  43. [51]

    ClassMix: Segmentation-based data augmentation for semi-supervised learning

    Viktor Olsson, Wilhelm Tranheden, Juliano Pinto, and Lennart Svensson. ClassMix: Segmentation-based data augmentation for semi-supervised learning. In WACV,

  44. [52]

    pix2gestalt: Amodal segmentation by synthesizing wholes

    Ege Ozguroglu, Ruoshi Liu, D´ıdac Sur´ıs, Dian Chen, Achal Dave, Pavel Tokmakov, and Carl V ondrick. pix2gestalt: Amodal segmentation by synthesizing wholes. In CVPR,

  45. [53]

    Seamless scene segmentation

    Lorenzo Porzi, Samuel Rota Bulo, Aleksander Colovic, and Peter Kontschieder. Seamless scene segmentation. In CVPR, 2019. 2

    2019

  46. [54]

    Amodal instance segmentation with KINS dataset

    Lu Qi, Li Jiang, Shu Liu, Xiaoyong Shen, and Jiaya Jia. Amodal instance segmentation with KINS dataset. In CVPR, 2019. 2

    2019

  47. [55]

    Source-free domain adaptation for RGB-D semantic seg- mentation with vision transformers

    Giulia Rizzoli, Donald Shenaj, and Pietro Zanuttigh. Source-free domain adaptation for RGB-D semantic seg- mentation with vision transformers. In WACV, 2024. 3

    2024

  48. [56]

    EDAPS: Enhanced domain-adaptive panoptic segmentation

    Suman Saha, Lukas Hoyer, Anton Obukhov, Dengxin Dai, and Luc Van Gool. EDAPS: Enhanced domain-adaptive panoptic segmentation. In ICCV, 2023. 3, 6, 7, 8, 14, 16, 17, 18

    2023

  49. [57]

    The omniscape dataset

    Ahmed Rida Sekkat, Yohan Dupuis, Pascal Vasseur, and Paul Honeine. The omniscape dataset. In ICRA, 2020. 2

    2020

  50. [58]

    AmodalSynthDrive: A syn- thetic amodal perception dataset for autonomous driving

    Ahmed Rida Sekkat, Rohit Mohan, Oliver Sawade, Elmar Matthes, and Abhinav Valada. AmodalSynthDrive: A syn- thetic amodal perception dataset for autonomous driving. IEEE Robotics and Automation Letters, 2024. 6, 13

    2024

  51. [59]

    Mo- tion parallax for 360° RGBD video

    Ana Serrano, Incheol Kim, Zhili Chen, Stephen DiVerdi, Diego Gutierrez, Aaron Hertzmann, and Belen Masia. Mo- tion parallax for 360° RGBD video. IEEE Transactions on Visualization and Computer Graphics, 2019. 1

    2019

  52. [60]

    FishDreamer: Towards fisheye seman- tic completion via unified image outpainting and segmenta- tion

    Hao Shi, Yu Li, Kailun Yang, Jiaming Zhang, Kunyu Peng, Alina Roitberg, Yaozu Ye, Huajian Ni, Kaiwei Wang, and Rainer Stiefelhagen. FishDreamer: Towards fisheye seman- tic completion via unified image outpainting and segmenta- tion. In CVPRW, 2023. 2

    2023

  53. [61]

    PanoFlow: Learning 360° optical flow for surrounding temporal understanding

    Hao Shi, Yifan Zhou, Kailun Yang, Xiaoting Yin, Ze Wang, Yaozu Ye, Zhe Yin, Shi Meng, Peng Li, and Kaiwei Wang. PanoFlow: Learning 360° optical flow for surrounding temporal understanding. IEEE Transactions on Intelligent Transportation Systems, 2023. 1

    2023

  54. [62]

    Amodal segmentation through out-of-task and out-of-distribution generalization with a Bayesian model

    Yihong Sun, Adam Kortylewski, and Alan Yuille. Amodal segmentation through out-of-task and out-of-distribution generalization with a Bayesian model. In CVPR, 2022. 2

    2022

  55. [63]

    Onur Tasar, S. L. Happy, Yuliya Tarabalka, and Pierre Al- liez. ColorMapGAN: Unsupervised domain adaptation for semantic segmentation using color mapping generative ad- versarial networks. IEEE Transactions on Geoscience and Remote Sensing, 2020. 3

    2020

  56. [64]

    AISFormer: Amodal instance segmentation with transformer

    Minh Tran, Khoa V o, Kashu Yamazaki, Arthur Fernandes, Michael Kidd, and Ngan Le. AISFormer: Amodal instance segmentation with transformer. In BMVC, 2022. 2

    2022

  57. [65]

    Amodal in- stance segmentation with diffusion shape prior estimation

    Minh Tran, Khoa V o, Tri Nguyen, and Ngan Le. Amodal in- stance segmentation with diffusion shape prior estimation. In ACCV, 2024. 3

    2024

  58. [66]

    DACS: Domain adaptation via cross- domain mixed sampling

    Wilhelm Tranheden, Viktor Olsson, Juliano Pinto, and Lennart Svensson. DACS: Domain adaptation via cross- domain mixed sampling. In WACV, 2021. 2

    2021

  59. [67]

    Visualizing data using t-SNE

    Laurens van der Maaten and Geoffrey Hinton. Visualizing data using t-SNE. Journal of Machine Learning Research,

  60. [68]

    FVP: Fourier visual prompting for source-free unsupervised do- main adaptation of medical image segmentation

    Yan Wang, Jian Cheng, Yixin Chen, Shuai Shao, Lanyun Zhu, Zhenzhou Wu, Tao Liu, and Haogang Zhu. FVP: Fourier visual prompting for source-free unsupervised do- main adaptation of medical image segmentation. IEEE Transactions on Medical Imaging, 2023. 3

    2023

  61. [69]

    Huang, and Honghui Shi

    Zhonghao Wang, Mo Yu, Yunchao Wei, Rog´erio Feris, Jin- jun Xiong, Wen-Mei Hwu, Thomas S. Huang, and Honghui Shi. Differential treatment for stuff and things: A simple unsupervised domain adaptation method for semantic seg- mentation. In CVPR, 2020. 3

    2020

  62. [70]

    Amodal segmentation based on vis- ible region segmentation and shape prior

    Yuting Xiao, Yanyu Xu, Ziming Zhong, Weixin Luo, Jiawei Li, and Shenghua Gao. Amodal segmentation based on vis- ible region segmentation and shape prior. In AAAI, 2021. 2

    2021

  63. [71]

    ´Alvarez, and Ping Luo

    Enze Xie, Wenhai Wang, Zhiding Yu, Anima Anandkumar, Jos´e M. ´Alvarez, and Ping Luo. SegFormer: Simple and ef- ficient design for semantic segmentation with transformers. In NeurIPS, 2021. 1

    2021

  64. [72]

    Neutral cross-entropy loss based un- supervised domain adaptation for semantic segmentation

    Hanqing Xu, Ming Yang, Liuyuan Deng, Yeqiang Qian, and Chunxiang Wang. Neutral cross-entropy loss based un- supervised domain adaptation for semantic segmentation. IEEE Transactions on Image Processing, 2021. 3

    2021

  65. [73]

    Amodal completion via progressive mixed context diffusion

    Katherine Xu, Lingzhi Zhang, and Jianbo Shi. Amodal completion via progressive mixed context diffusion. In CVPR, 2024. 3

    2024

  66. [74]

    Source free domain adaptation for semantic segmentation via distribution transfer and adaptive class-balanced self- training

    Cheng-Yu Yang, Yuan-Jhe Kuo, and Chiou-Ting Hsu. Source free domain adaptation for semantic segmentation via distribution transfer and adaptive class-balanced self- training. In ICME, 2022. 8

    2022

  67. [75]

    Can we PASS beyond the field of view? Panoramic annular semantic segmentation for real-world surrounding percep- tion

    Kailun Yang, Xinxin Hu, Luis Miguel Bergasa, Eduardo Romera, Xiao Huang, Dongming Sun, and Kaiwei Wang. Can we PASS beyond the field of view? Panoramic annular semantic segmentation for real-world surrounding percep- tion. In IV, 2019. 2

    2019

  68. [76]

    PASS: Panoramic annular semantic segmentation

    Kailun Yang, Xinxin Hu, Luis Miguel Bergasa, Eduardo Romera, and Kaiwei Wang. PASS: Panoramic annular semantic segmentation. IEEE Transactions on Intelligent Transportation Systems, 2020. 2

    2020

  69. [77]

    Is context-aware CNN ready for the surroundings? Panoramic semantic segmentation in the wild

    Kailun Yang, Xinxin Hu, and Rainer Stiefelhagen. Is context-aware CNN ready for the surroundings? Panoramic semantic segmentation in the wild. IEEE Transactions on Image Processing, 2021. 2

    2021

  70. [78]

    Capturing omni-range context for om- nidirectional segmentation

    Kailun Yang, Jiaming Zhang, Simon Reiß, Xinxin Hu, and Rainer Stiefelhagen. Capturing omni-range context for om- nidirectional segmentation. In CVPR, 2021. 1

    2021

  71. [79]

    Universal semantic segmentation for fisheye ur- ban driving images

    Yaozu Ye, Kailun Yang, Kaite Xiang, Juan Wang, and Kai- wei Wang. Universal semantic segmentation for fisheye ur- ban driving images. In SMC, 2020. 2

    2020

  72. [80]

    When source-free domain adaptation meets learning with noisy labels

    Li Yi, Gezheng Xu, Pengcheng Xu, Jiaqi Li, Ruizhi Pu, Charles Ling, Allan Mcleod, and Boyu Wang. When source-free domain adaptation meets learning with noisy labels. In ICLR, 2023. 3

    2023

  73. [81]

    WoodScape: A multi-task, multi-camera fisheye dataset for autonomous driving

    Senthil Kumar Yogamani, Christian Witt, Hazem Rashed, Sanjaya Nayak, Saquib Mansoor, Padraig Varley, Xavier Perrotton, Derek O’Dea, Patrick P ´erez, Ciar ´an Hughes, Jonathan Horgan, Ganesh Sistu, Sumanth Chennupati, Michal Uric ´ar, Stefan Milz, Martin Simon, and Karl Amende....

    2019

  74. [82]

    A comprehensive survey on source-free domain adaptation

    Zhiqi Yu, Jingjing Li, Zhekai Du, Lei Zhu, and Heng Tao Shen. A comprehensive survey on source-free domain adaptation. IEEE Transactions on Pattern Analysis and Ma- chine Intelligence, 2024. 3

    2024

  75. [83]

    Dauphin, and David Lopez-Paz

    Hongyi Zhang, Moustapha Ciss ´e, Yann N. Dauphin, and David Lopez-Paz. mixup: Beyond empirical risk minimiza- tion. In ICLR, 2018. 5

    2018

  76. [84]

    IS- SAFE: Improving semantic segmentation in accidents by fusing event-based data

    Jiaming Zhang, Kailun Yang, and Rainer Stiefelhagen. IS- SAFE: Improving semantic segmentation in accidents by fusing event-based data. In IROS, 2021. 2

    2021

  77. [85]

    Transfer be- yond the field of view: Dense panoramic semantic segmen- tation via unsupervised domain adaptation

    Jiaming Zhang, Chaoxiang Ma, Kailun Yang, Alina Roit- berg, Kunyu Peng, and Rainer Stiefelhagen. Transfer be- yond the field of view: Dense panoramic semantic segmen- tation via unsupervised domain adaptation. IEEE Transac- tions on Intelligent Transportation Systems, 2022. 2, 3

    2022

  78. [86]

    Bending reality: Distortion-aware transformers for adapting to panoramic semantic segmentation

    Jiaming Zhang, Kailun Yang, Chaoxiang Ma, Simon Reiß, Kunyu Peng, and Rainer Stiefelhagen. Bending reality: Distortion-aware transformers for adapting to panoramic semantic segmentation. In CVPR, 2022. 2, 3, 6, 7, 8, 14, 16, 17, 18

    2022

  79. [87]

    Hierarchi- cal mask calibration for unified domain adaptive panoptic segmentation

    Jingyi Zhang, Jiaxing Huang, and Shijian Lu. Hierarchi- cal mask calibration for unified domain adaptive panoptic segmentation. In CVPR, 2023. 6, 7, 8, 18

    2023

  80. [88]

    UniDAformer: Unified domain adaptive panoptic seg- mentation transformer via hierarchical mask calibration

    Jingyi Zhang, Jiaxing Huang, Xiaoqin Zhang, and Shijian Lu. UniDAformer: Unified domain adaptive panoptic seg- mentation transformer via hierarchical mask calibration. In CVPR, 2023. 3

    2023

  81. [89]

    Jiaming Zhang, Kailun Yang, Hao Shi, Simon Reiß, Kunyu Peng, Chaoxiang Ma, Haodong Fu, Philip H. S. Torr, Kai- wei Wang, and Rainer Stiefelhagen. Behind every domain there is a shift: Adapting distortion-aware vision transform- ers for panoramic semantic segmentation. IEEE Tra...

    2024

  82. [90]

    GoodSAM: Bridging domain and capacity gaps via seg- ment anything model for distortion-aware panoramic se- mantic segmentation

    Weiming Zhang, Yexin Liu, Xu Zheng, and Lin Wang. GoodSAM: Bridging domain and capacity gaps via seg- ment anything model for distortion-aware panoramic se- mantic segmentation. In CVPR, 2024

    2024

  83. [91]

    GoodSAM++: Bridging domain and capacity gaps via seg- ment anything model for panoramic semantic segmenta- tion

    Weiming Zhang, Yexin Liu, Xu Zheng, and Lin Wang. GoodSAM++: Bridging domain and capacity gaps via seg- ment anything model for panoramic semantic segmenta- tion. arXiv preprint arXiv:2408.09115, 2024. 3

    2024 arXiv

  84. [92]

    Un- supervised domain adaptation for semantic segmentation with pseudo label self-refinement

    Xingchen Zhao, Niluthpol Chowdhury Mithun, Abhinav Rajvanshi, Han-Pang Chiu, and Supun Samarasekera. Un- supervised domain adaptation for semantic segmentation with pseudo label self-refinement. In WACV, 2023. 3

    2023

  85. [93]

    Open panoramic segmentation

    Junwei Zheng, Ruiping Liu, Yufan Chen, Kunyu Peng, Chengzhi Wu, Kailun Yang, Jiaming Zhang, and Rainer Stiefelhagen. Open panoramic segmentation. In ECCV,

  86. [94]

    Look at the neighbor: Distortion-aware unsupervised domain adaptation for panoramic semantic segmentation

    Xu Zheng, Tianbo Pan, Yunhao Luo, and Lin Wang. Look at the neighbor: Distortion-aware unsupervised domain adaptation for panoramic semantic segmentation. In ICCV,

  87. [95]

    Both style and distortion matter: Dual- path unsupervised domain adaptation for panoramic seman- tic segmentation

    Xu Zheng, Jinjing Zhu, Yexin Liu, Zidong Cao, Chong Fu, and Lin Wang. Both style and distortion matter: Dual- path unsupervised domain adaptation for panoramic seman- tic segmentation. In CVPR, 2023. 3

    2023

  88. [96]

    Vasilakos, and Lin Wang

    Xu Zheng, Pengyuan Zhou, Athanasios V . Vasilakos, and Lin Wang. 360SFUDA++: Towards source-free UDA for panoramic segmentation by learning reliable category pro- totypes. IEEE Transactions on Pattern Analysis and Ma- chine Intelligence, 2024. 2, 3, 6, 7, 8, 14, 17, 18

    2024

  89. [97]

    Vasilakos, and Lin Wang

    Xu Zheng, Pengyuan Zhou, Athanasios V . Vasilakos, and Lin Wang. Semantics distortion and style matter: Towards source-free UDA for panoramic segmentation. In CVPR,

  90. [98]

    Complementary bi-directional feature compression for indoor 360° semantic segmentation with self-distillation

    Zishuo Zheng, Chunyu Lin, Lang Nie, Kang Liao, Zhijie Shen, and Yao Zhao. Complementary bi-directional feature compression for indoor 360° semantic segmentation with self-distillation. In WACV, 2023. 2

    2023

  91. [99]

    Semantic amodal segmentation

    Yan Zhu, Yuandong Tian, Dimitris Metaxas, and Piotr Doll´ar. Semantic amodal segmentation. In CVPR, 2017. 2

    2017

  92. [100]

    Yang Zou, Zhiding Yu, B. V . K. Vijaya Kumar, and Jinsong Wang. Unsupervised domain adaptation for semantic seg- mentation via class-balanced self-training. In ECCV, 2018. 3

    2018

  93. [102]

    As depicted in Fig

    SFOASS vs OASS Occlusion-Aware Seamless Segmentation (OASS), intro- duced by [7], aims to address the issues of the narrow field of view, occlusion of perspective, and domain gaps in a seamless manner. As depicted in Fig. 7a, OASS requires the model to transfer from a labeled ...

  94. [103]

    These benchmarks are based on three datasets: KITTI360-APS [49], AmodalSynthDrive [58], and BlendPASS [7]

    Benchmarks In this work, we applied two benchmarks to evaluate the SFOASS task. These benchmarks are based on three datasets: KITTI360-APS [49], AmodalSynthDrive [58], and BlendPASS [7]. In accordance with the SFOASS task formulation, both KITTI360-APS and AmodalSyn- thDrive s...

  95. [104]

    This cri- terion ensures that excessive contextual information about the objects is not discarded under the spatial-aware mix- ing strategy

    More Experiments Details In the Amodal-Driven Contextual Learning (ADCL) of UN- LOCK, only object samples with an overlapping area of the object region that is less than half of the full area of the object are stored in the amodal-driven object pool. This cri- terion ensures t...

  96. [105]

    We specifically analyze the effect of the hyperparameters in both OPLL and ADCL on the KITTI360-APS →BlendPASS benchmark

    Analysis for Hyper-parameters In UNLOCK, OPLL serves as the key, and ADCL per- forms the unlocking action. We specifically analyze the effect of the hyperparameters in both OPLL and ADCL on the KITTI360-APS →BlendPASS benchmark. For the SFOASS task, performance is evaluated us...

  97. [106]

    More Visualization Results on SFOASS As shown in Fig

    More Visualization Results 10.1. More Visualization Results on SFOASS As shown in Fig. 9, we further compare our proposed UNLOCK with existing UDA methods [7, 56, 86, 94] on KITTI360-APS →BlendPASS benchmark. The results demonstrate that, even without access to source domain i...

  98. [107]

    Discussion 11.1. Limitations and potential solutions The 360° boundaries of panoramic images remain an under- explored aspect in the field of seamless segmentation, pos- ing challenges for accurate perception at extreme viewing angles. Additionally, the interaction between the...

  99. [2023]

    3, 6, 7, 8, 14, 16, 17, 18

Pith tools

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