Pith. sign in

REVIEW 3 major objections 4 minor 1 cited by

NVSPolicy: Adaptive Novel-View Synthesis for Generalizable Language-Conditioned Policy Learning

T0 review · 3 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read Adaptive novel-view synthesis lifts language-conditioned robot success to 90.4%.

desk verdict A plausible integration paper whose adaptive-viewpoint novelty is not statistically supported; deserves revision, not rejection. read the letter →

arxiv 2505.10359 v1 pith:P75ARCLE submitted 2025-05-15 cs.RO cs.CV

Le Shi , Yifei Shi , Xin Xu , Tenglong Liu , Junhua Xi , Chengyuan Chen This is my paper · ORCID
classification cs.ROcs.CV
keywords novel-viewsynthesislanguage-conditionedpolicylearninghierarchicalcycle-consistentVAEfeaturedisentanglementCALVINbenchmarkimitationrobotmanipulation
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

NVSPolicy claims that a robot following natural-language instructions can be made more reliable by first generating an extra camera view of the scene from a viewpoint chosen on the fly, then feeding that synthetic view through a two-part encoder. The semantic part of the encoded features chooses which high-level skill to execute, while the remaining part predicts low-level motion, so flaws in the generated image do not corrupt the skill choice. On the CALVIN long-horizon benchmark the method reaches an average success rate of 90.4% and completes 2.93 consecutive tasks on average, above the best prior methods that do not use large foundation models, and it also beats a baseline on a physical robot. The point of the work is that imperfect generative images are usable for control if the policy separates what is in the scene from everything else.

What carries the argument

The load-bearing object is the adaptive viewpoint rule $\theta = -w_1 d + w_2$ with $w_1=14$ and $w_2=39$, where $d$ is the average scene depth; this sets how far the synthesized camera swings away from the observed view. GenWarp, a pre-trained single-image generative warping model, then renders that novel view. Around this sit a cycle-consistent VAE that disentangles each image into semantic and remaining latent features, a hierarchical policy in which the semantic features drive meta-skill selection and the remaining features drive the action estimator, and two efficiency mechanisms, keyframe-only synthesis and policy distillation, that let a student network estimate semantic features without running the generative model at every step.

What would settle it

Move the wrist camera to a different mounting height on the same simulator and rerun the identical pipeline with $w_1=14$, $w_2=39$: if the average horizon length is not above the best fixed angle, roughly 20 degrees, or if a grid search over the coefficients shows a broad plateau, then the adaptive viewpoint rule is not doing the causal work claimed.

Watch

Extended reading notes

Core claim

The paper's central claim is that synthesizing a novel viewpoint from the wrist camera, with the viewpoint chosen by a simple depth-based rule, gives a language-conditioned policy enough extra visual context to generalize better across long task chains. To make the imperfect synthetic image safe, the authors train a cycle-consistent VAE that splits each image into a semantic feature, shared across views and used for meta-skill selection, and a remaining feature, used for action estimation. Ablations support the claim: removing novel-view synthesis lowers the average horizon length from 2.93 to 2.71, removing feature disentanglement lowers it to 2.73, and replacing the adaptive angle with the best fixed angle of 20 degrees lowers it to 2.89. The authors conclude that adaptive novel-view synthesis plus disentangled features is the source of the improvement.

Load-bearing premise

The load-bearing premise is that the hand-set linear viewpoint rule $\theta = -w_1 d + w_2$, with $w_1=14$ and $w_2=39$, generalizes beyond CALVIN's camera setup; it is an ad hoc heuristic with no derivation, and a fixed 20-degree viewpoint nearly matches its average horizon length, 2.89 versus 2.93, so the adaptive advantage rests on coefficients that could be benchmark-specific.

Editorial extensions

If this is right

  • If the central claim holds, a robot policy can safely consume generative images: semantic features are robust to synthetic artifacts while action-relevant details come from real cameras.
  • Removing the synthesized view costs about 0.22 tasks of average horizon length, so the novel-view context is doing real work for long task chains.
  • Adaptive viewpoint selection outperforms every fixed angle tested, with the largest margin at horizon 5, 32.8% versus 29.2% for 20 degrees.
  • Policy distillation can replace explicit synthesis during inference, meaning the generative model acts as a training-time teacher rather than a runtime cost.
  • The same architecture transfers to a physical robot, with average success 44% versus 8% for the compared baseline across five tasks.

Reading between the lines

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

  • A natural next test is whether the viewpoint rule $\theta = -14 d + 39$ transfers: on a benchmark with a different camera rig or depth distribution the same coefficients may need re-tuning, and a grid search would show whether the adaptive advantage is a sharp optimum or a broad plateau.
  • Because the policy deliberately ignores the synthesized image's remaining feature, the same semantic-versus-remaining disentanglement could admit other imperfect generative inputs, such as inpainted occlusions or hallucinated goal states, without retraining the action estimator.
  • Keyframe selection plus distillation implies the synthesized view acts mainly as an auxiliary training signal, which suggests a fully self-supervised variant could train the student directly on sampled viewpoints and remove the teacher at deployment.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 4 minor

Summary. The paper proposes NVSPolicy, a language-conditioned robotic manipulation policy that augments the gripper-camera observation with a novel-view image synthesized by GenWarp at an adaptively selected viewpoint. A cycle-consistent VAE disentangles the visual features into a semantic feature and a remaining feature; the semantic feature drives a high-level meta-skill selector, while the remaining feature drives a low-level action estimator. The method also includes keyframe selection and policy distillation for efficiency. The paper reports state-of-the-art results on CALVIN (Horizon-1 success 90.4%, average horizon length 2.93), ablation studies isolating the contribution of novel-view synthesis, feature disentanglement, and policy distillation, and a small real-robot comparison against SPIL.

Significance. If the reported gains are statistically robust, the paper makes a useful contribution: it demonstrates that a pretrained single-image novel-view synthesizer can provide task-relevant context for language-conditioned policy learning, and that disentangling semantic from remaining features is a sensible way to mitigate synthesis artifacts. Evaluation on the public CALVIN benchmark against several recent non-foundation-model baselines is a strength, as is the inclusion of ablation and real-robot studies. However, the statistical support for the key adaptive-viewpoint mechanism is currently weak, and an internal inconsistency in Table III limits what can be concluded about the feature-disentanglement component.

major comments (3)
  1. [Table III / Sec. IV-D] The rows 'w/o NS' and 'w/o FD' report identical success rates at Horizons 1, 2, 3, and 5 (88.3/88.3, 68.7/68.7, 51.3/51.3, 26.1/26.1) but different average horizon lengths (2.71 vs. 2.73), and the text states a reduction to 2.74 for w/o FD. This internal inconsistency makes it impossible to evaluate the contribution of feature disentanglement, which is a load-bearing component of the proposed method. Please correct the table and text, and either rerun the w/o FD ablation or report the actual numbers with the correct averages.
  2. [Table II / Sec. IV-D] The adaptive viewpoint rule is the central novelty of the paper, but Table II shows that the fixed angle θ=20° achieves an average horizon length of 2.89 ± 0.06 versus 2.93 ± 0.04 for the adaptive rule, and at Horizon 2 the fixed angle is higher (71.2 vs. 70.4). No significance test, paired comparison, or per-seed breakdown is reported, so the evidence that adaptive selection outperforms a fixed viewpoint is not demonstrated. Please provide statistical tests (e.g., paired tests across seeds or task sequences) and report per-seed results, or temper the claim accordingly.
  3. [Sec. III-B, Eq. (1)] The linear heuristic θ = −w1·d_cam-ori + w2 with w1=14 and w2=39 is introduced without derivation, without specifying the units or valid range of d_cam-ori, and without sensitivity analysis. The 'small location perturbation' added to the sampled viewpoint is also unspecified. If the coefficients were tuned on CALVIN's depth statistics, the generalization claim beyond CALVIN is not supported. Please justify the heuristic, report the depth range over which it was designed or tuned, and include a sensitivity study over w1, w2, and the perturbation magnitude.
minor comments (4)
  1. [Sec. IV-E / Table IV] The real-robot evaluation uses only 10 trials per task and reports a single baseline (SPIL) with no variance, error bars, or statistical test; please state this limitation explicitly and, if possible, report per-trial outcomes or confidence intervals.
  2. [Sec. I, Sec. III-C, Sec. III-D] There are several typographical and grammatical errors, e.g., 'can generation effort' in Sec. I, 'a adaptive novel viewpoint' in Sec. I, 'mechansim' in Sec. III-C, and 'we does not consider' in Sec. III-D; a careful proofread is needed.
  3. [Appendix VI-B] Several symbols in the notation list (MI(·), LKD, m, n, τ, K, p(It−1, It)) are not used in the main text, which appears to be leftover material from an earlier draft; please remove them or introduce them in the appropriate sections.
  4. [Figs. 2 and 4] The arrows and labels in Figs. 2 and 4 are dense and partly illegible, especially the feature-exchange paths in the forward and reverse processes of the cycle-consistent VAE; please enlarge fonts or add a clearer explanatory caption.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the central components are trained on demonstration data or adopted from external prior work, and the adaptive-angle heuristic is an empirical design choice, not a prediction derived from the success metric.

full rationale

The paper's derivation chain is self-contained against external benchmarks. The novel-view synthesis module relies on GenWarp [8], an externally published generative model; the cycle-consistent VAE is trained on image pairs with synthetic distortions, not on task success; the hierarchical policy is trained on demonstration actions with rule-based meta-skill labels derived from actions via SPIL [6]. Equation (1), the adaptive viewpoint rule, is a manually specified linear heuristic with hyperparameters w1=14 and w2=39, but the paper does not present it as a fitted parameter that predicts a held-out quantity, nor does any equation reduce to the success metric. The adaptive-vs-fixed-angle comparison in Table II is an empirical ablation on the public CALVIN benchmark; its statistical weakness and possible sensitivity to hyperparameter tuning are correctness risks, not circularity. The ablation table shows nearly identical w/o NS and w/o FD rows, which may indicate a copy error, but this again affects experimental reporting, not logical circularity. There are no self-citations that carry the argument, no imported uniqueness theorems, and no renaming of known results as new predictions. Therefore, no circular step is identifiable under the specified criteria.

Assumptions & free parameters 6 free parameters · 7 assumptions · 0 invented entities

The central method rests on several hand-chosen hyperparameters and external pretrained components. The most important fitted parameters are the coefficients of the adaptive viewpoint rule, w1 and w2. The cycle-consistent VAE transfer assumption is unverified, and no new physical or conceptual entities are introduced.

free parameters (6)
  • w1, w2 in viewpoint angle rule = w1=14, w2=39
    Eq. (1) maps scene depth to viewpoint offset; the coefficients are set by hand with no derivation and are not tested on other distributions.
  • lambda (cycle-VAE balance) = 0.5
    Balances forward and reverse losses of the cycle-consistent VAE; chosen by hand.
  • gamma1, gamma2 (hierarchical loss weights) = gamma1=0.2, gamma2=0.01
    Weights for meta-skill cross-entropy and KL regularization in Eq. (4); chosen by hand.
  • T (rolling buffer length) = 15
    Number of past semantic features kept for meta-skill selection; chosen by hand.
  • Keyframe angle threshold = not reported
    Keyframes are selected when viewpoint change exceeds a threshold, but the threshold value is never given; needed to reproduce training.
  • Viewpoint perturbation magnitude = not reported
    A 'small location perturbation' is added to the sampled viewpoint in Section III-B without specifying its distribution or scale.
assumptions (7)
  • domain assumption GenWarp produces semantically coherent novel views from a single image
    The entire policy relies on the synthesized image having correct semantics; GenWarp is a frozen external model with known failure modes under large viewpoint shifts, which the authors acknowledge.
  • domain assumption Average scene depth estimated by the method in [8] is accurate enough for viewpoint selection
    d_cam_ori is set equal to this estimated average depth and directly determines theta in Eq. (1).
  • domain assumption Cycle-consistent VAE trained on same-scene image pairs with random distortions generalizes to real GenWarp artifacts
    The training pairs are augmented with local geometric distortions and appearance changes to mimic GenWarp outputs, but no evidence is given that these match actual GenWarp artifacts.
  • domain assumption Rule-based meta-skill labels p(o|a) from SPIL are valid for CALVIN and the real robot
    The meta-skill selector is supervised with this external labeling rule; errors in the rule propagate to skill selection and action estimation.
  • domain assumption CALVIN's standardized initialization and unspecified split measures generalizable language-conditioned control
    The paper reports CALVIN results without specifying the training/evaluation split or confirming that the numbers match the public leaderboard protocol.
  • ad hoc to paper The linear depth-angle relation in Eq. (1) is an appropriate heuristic for balancing information gain and synthesis fidelity
    No derivation or theoretical justification is provided; the coefficients are tuned, and the relation is specific to this paper.
  • domain assumption MiniLM language embeddings are sufficient to ground instructions
    Language conditioning uses a pretrained MiniLM encoder with no adaptation; if instruction grounding is poor, the hierarchical policy degrades.

how reviews work

0 comments
Cite this review

Pith. "Pith review of NVSPolicy: Adaptive Novel-View Synthesis for Generalizable Language-Conditioned Policy Learning." pith.science (2026). https://pith.science/paper/P75ARCLE

@misc{pith2026250510359,
  author       = {Pith},
  title        = {Pith review of: NVSPolicy: Adaptive Novel-View Synthesis for Generalizable Language-Conditioned Policy Learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/P75ARCLE}},
  note         = {Machine review of arXiv:2505.10359}
}
read the original abstract

Recent advances in deep generative models demonstrate unprecedented zero-shot generalization capabilities, offering great potential for robot manipulation in unstructured environments. Given a partial observation of a scene, deep generative models could generate the unseen regions and therefore provide more context, which enhances the capability of robots to generalize across unseen environments. However, due to the visual artifacts in generated images and inefficient integration of multi-modal features in policy learning, this direction remains an open challenge. We introduce NVSPolicy, a generalizable language-conditioned policy learning method that couples an adaptive novel-view synthesis module with a hierarchical policy network. Given an input image, NVSPolicy dynamically selects an informative viewpoint and synthesizes an adaptive novel-view image to enrich the visual context. To mitigate the impact of the imperfect synthesized images, we adopt a cycle-consistent VAE mechanism that disentangles the visual features into the semantic feature and the remaining feature. The two features are then fed into the hierarchical policy network respectively: the semantic feature informs the high-level meta-skill selection, and the remaining feature guides low-level action estimation. Moreover, we propose several practical mechanisms to make the proposed method efficient. Extensive experiments on CALVIN demonstrate the state-of-the-art performance of our method. Specifically, it achieves an average success rate of 90.4\% across all tasks, greatly outperforming the recent methods. Ablation studies confirm the significance of our adaptive novel-view synthesis paradigm. In addition, we evaluate NVSPolicy on a real-world robotic platform to demonstrate its practical applicability.

Figures

Figures reproduced from arXiv: 2505.10359 by the authors.

Figure 1
Figure 1. The proposed NVSPolicy leverages adaptive novel [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The overview of the proposed NVSPolicy. (a) It synthesizes context-enhanced novel-view images given the input [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. (a) The local spherical coordinate in the adaptive [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: The training protocol of the cycle-consistent VAE: a forward process and a reverse process are developed to jointly [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]
Figure 5
Figure 5. Figure 5: Qualitative comparisons of NVSPolicy with several recent methods. Frames marked with a check mark indicate a [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 6
Figure 6. Figure 6: The proposed NVSPolicy shows satisfactory perfor [PITH_FULL_IMAGE:figures/full_fig_p007_6.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. Global Pose Control for Generative View Synthesis in Normalized Object Coordinate Space

    cs.CV 2026-07 conditional novelty 6.5 of 10

    A diffusion image-editing model conditioned on Plücker ray-map tokens and text-defined NOCS fronts generates high-fidelity novel views with absolute global pose control from unposed inputs.

Reference graph

Works this paper leans on

38 extracted references · 32 canonical work pages · cited by 1 Pith paper

  1. [1]

    Large language models encode clinical knowledge

    Karan Singhal, Shekoofeh Azizi, Tao Tu, S Sara Mahdavi, Jason Wei, Hyung Won Chung, Nathan Scales, Ajay Tanwani, Heather Cole- Lewis, Stephen Pfohl, et al. Large language models encode clinical knowledge. Nature, 620(7972):172–180, 2023

    work page 2023

  2. [2]

    Palm: Scaling language modeling with pathways

    Aakanksha Chowdhery, Sharan Narang, Jacob Devlin, Maarten Bosma, Gaurav Mishra, Adam Roberts, Paul Barham, Hyung Won Chung, Charles Sutton, Sebastian Gehrmann, et al. Palm: Scaling language modeling with pathways. Journal of Machine Learning Research, 24(240):1–113, 2023

    2023

  3. [3]

    3d-mvp: 3d multiview pretraining for robotic manipulation

    Shengyi Qian, Kaichun Mo, Valts Blukis, David F Fouhey, Dieter Fox, and Ankit Goyal. 3d-mvp: 3d multiview pretraining for robotic manipulation. arXiv preprint arXiv:2406.18158 , 2024

    arXiv 2024

  4. [4]

    3d diffusion policy: Generalizable visuomotor policy learning via simple 3d representations

    Yanjie Ze, Gu Zhang, Kangning Zhang, Chenyuan Hu, Muhan Wang, and Huazhe Xu. 3d diffusion policy: Generalizable visuomotor policy learning via simple 3d representations. In International Conference on Robotics and Automation (ICRA) , 2024

    work page 2024

  5. [5]

    Vima: robot manipulation with multimodal prompts

    Yunfan Jiang, Agrim Gupta, Zichen Zhang, Guanzhi Wang, Yongqiang Dou, Yanjun Chen, Li Fei-Fei, Anima Anandkumar, Yuke Zhu, and Linxi Fan. Vima: robot manipulation with multimodal prompts. In International Conference on Machine Learning (ICML) , pages 14975– 15022, 2023

    work page 2023

  6. [6]

    Language-conditioned imitation learning with base skill priors under unstructured data

    Hongkuan Zhou, Zhenshan Bing, Xiangtong Yao, Xiaojie Su, Chen- guang Yang, Kai Huang, and Alois Knoll. Language-conditioned imitation learning with base skill priors under unstructured data. IEEE Robotics and Automation Letters , 9(11):9805–9812, 2024

    work page 2024

  7. [7]

    What matters in language conditioned robotic imitation learning over unstructured data

    Oier Mees, Lukas Hermann, and Wolfram Burgard. What matters in language conditioned robotic imitation learning over unstructured data. IEEE Robotics and Automation Letters , 7(4):11205–11212, 2022

    2022

  8. [8]

    Genwarp: Single image to novel views with semantic- preserving generative warping

    Junyoung Seo, Kazumi Fukuda, Takashi Shibuya, Takuya Narihira, Naoki Murata, Shoukang Hu, Chieh-Hsin Lai, Seungryong Kim, and Yuki Mitsufuji. Genwarp: Single image to novel views with semantic- preserving generative warping. In Conference on Neural Information Processing Systems(NIPS), 2024

    work page 2024

Show all 38 references
  1. [9]

    Eschernet: A generative model for scalable view synthesis

    Xin Kong, Shikun Liu, Xiaoyang Lyu, Marwan Taher, Xiaojuan Qi, and Andrew J Davison. Eschernet: A generative model for scalable view synthesis. In Computer Vision and Pattern Recognition (CVPR) , pages 9503–9513, 2024

    2024

  2. [10]

    Integrated learning of robot motion and sentences: Real-time prediction of grasping motion and attention based on language instructions

    Hiroshi Ito, Hideyuki Ichiwara, Kenjiro Yamamoto, Hiroki Mori, and Tetsuya Ogata. Integrated learning of robot motion and sentences: Real-time prediction of grasping motion and attention based on language instructions. In International Conference on Robotics and Automation (IC...

    2022

  3. [11]

    A joint modeling of vision- language-action for target-oriented grasping in clutter

    Kechun Xu, Shuqi Zhao, Zhongxiang Zhou, Zizhang Li, Huaijin Pi, Yifeng Zhu, Yue Wang, and Rong Xiong. A joint modeling of vision- language-action for target-oriented grasping in clutter. In International Conference on Robotics and Automation (ICRA) , 2023

    2023

  4. [12]

    Alphablock: Embodied finetuning for vision-language reasoning in robot manipulation

    Chuhao Jin, Wenhui Tan, Jiange Yang, Bei Liu, Ruihua Song, Limin Wang, and Jianlong Fu. Alphablock: Embodied finetuning for vision-language reasoning in robot manipulation. arXiv preprint arXiv:2305.18898, 2023

    2023 arXiv

  5. [13]

    Rt-grasp: Reasoning tuning robotic grasping via multi-modal large language model

    Jinxuan Xu, Shiyu Jin, Yutian Lei, Yuqian Zhang, and Liangjun Zhang. Rt-grasp: Reasoning tuning robotic grasping via multi-modal large language model. In International Conference on Intelligent Robots and Systems (IROS) , pages 7323–7330. IEEE, 2024

    2024

  6. [14]

    Chain-of-thought prompting elicits reasoning in large language models

    Jason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma, Fei Xia, Ed Chi, Quoc V Le, Denny Zhou, et al. Chain-of-thought prompting elicits reasoning in large language models. Advances in Neural Information Processing Systems (NIPS) , 35:24824–24837, 2022

    2022

  7. [15]

    Chatgpt for robotics: Design principles and model abilities

    Sai H Vemprala, Rogerio Bonatti, Arthur Bucker, and Ashish Kapoor. Chatgpt for robotics: Design principles and model abilities. IEEE Access, 12:55682–55696, 2024

    2024

  8. [16]

    Language conditioned imitation learning over unstructured data

    Corey Lynch and Pierre Sermanet. Language conditioned imitation learning over unstructured data. In arXiv preprint arXiv:2005.07648 , 2020

    2005 arXiv

  9. [17]

    Learning hand-eye coordination for robotic grasping with deep learning and large-scale data collection

    Sergey Levine, Peter Pastor, Alex Krizhevsky, Julian Ibarz, and Deirdre Quillen. Learning hand-eye coordination for robotic grasping with deep learning and large-scale data collection. The International Journal of Robotics Research , 37(4-5):421–436, 2018

    2018

  10. [18]

    Scalable deep reinforcement learning for vision-based robotic manipulation

    Dmitry Kalashnikov, Alex Irpan, Peter Pastor, Julian Ibarz, Alexander Herzog, Eric Jang, Deirdre Quillen, Ethan Holly, Mrinal Kalakrishnan, Vincent Vanhoucke, et al. Scalable deep reinforcement learning for vision-based robotic manipulation. In Conference on Robot Learning (Co...

    2018

  11. [19]

    Planning with diffusion for flexible behavior synthesis

    Michael Janner, Yilun Du, Joshua Tenenbaum, and Sergey Levine. Planning with diffusion for flexible behavior synthesis. In Interna- tional Conference on Machine Learning (ICML) , pages 9902–9915. PMLR, 2022

    2022

  12. [20]

    Language control diffusion: Efficiently scaling through space, time, and tasks

    Edwin Zhang, Yujie Lu, Shinda Huang, William Yang Wang, and Amy Zhang. Language control diffusion: Efficiently scaling through space, time, and tasks. In International Conference on Learning Representations (ICLR), 2024

    2024

  13. [21]

    Rt- 2: Vision-language-action models transfer web knowledge to robotic control

    Brianna Zitkovich, Tianhe Yu, Sichun Xu, Peng Xu, Ted Xiao, Fei Xia, Jialin Wu, Paul Wohlhart, Stefan Welker, Ayzaan Wahid, et al. Rt- 2: Vision-language-action models transfer web knowledge to robotic control. In Conference on Robot Learning (CoRL) , pages 2165–2183. PMLR, 2023

    2023

  14. [22]

    Vision-language foundation models as effective robot imitators

    Xinghang Li, Minghuan Liu, Hanbo Zhang, Cunjun Yu, Jie Xu, Hongtao Wu, Chilam Cheang, Ya Jing, Weinan Zhang, Huaping Liu, et al. Vision-language foundation models as effective robot imitators. In International Conference on Learning Representations (ICLR) , 2024

    2024

  15. [23]

    Look before you leap: Unveiling the power of gpt-4v in robotic vision- language planning

    Yingdong Hu, Fanqi Lin, Tong Zhang, Li Yi, and Yang Gao. Look before you leap: Unveiling the power of gpt-4v in robotic vision- language planning. In International Conference on Robotics and Automation (ICRA) , 2024

    2024

  16. [24]

    Nerf: Representing scenes as neural radiance fields for view synthesis

    Ben Mildenhall, Pratul P Srinivasan, Matthew Tancik, Jonathan T Barron, Ravi Ramamoorthi, and Ren Ng. Nerf: Representing scenes as neural radiance fields for view synthesis. In European Conference on Computer Vision (ECCV) , pages 405–421, 2020

    2020

  17. [25]

    Instant neural graphics primitives with a multiresolution hash encod- ing

    Thomas M ¨uller, Alex Evans, Christoph Schied, and Alexander Keller. Instant neural graphics primitives with a multiresolution hash encod- ing. ACM Transactions on Graphics , 41(4):1–15, 2022

    2022

  18. [26]

    Uv volumes for real-time rendering of editable free-view human performance

    Yue Chen, Xuan Wang, Xingyu Chen, Qi Zhang, Xiaoyu Li, Yu Guo, Jue Wang, and Fei Wang. Uv volumes for real-time rendering of editable free-view human performance. In Conference on Computer Vision and Pattern Recognition (CVPR) , pages 16621–16631, 2023

    2023

  19. [27]

    3d gaussian splatting for real-time radiance field rendering

    Bernhard Kerbl, Georgios Kopanas, Thomas Leimk ¨uhler, and George Drettakis. 3d gaussian splatting for real-time radiance field rendering. ACM Trans. Graph. , 42(4):139–1, 2023

    2023

  20. [28]

    Compressed 3d gaussian splatting for accelerated novel view synthesis

    Simon Niedermayr, Josef Stumpfegger, and R ¨udiger Westermann. Compressed 3d gaussian splatting for accelerated novel view synthesis. In Computer Vision and Pattern Recognition (CVPR) , pages 10349– 10358, 2024

    2024

  21. [29]

    pi-gan: Periodic implicit generative adversarial networks for 3d-aware image synthesis

    Eric R Chan, Marco Monteiro, Petr Kellnhofer, Jiajun Wu, and Gordon Wetzstein. pi-gan: Periodic implicit generative adversarial networks for 3d-aware image synthesis. In Computer Vision and Pattern Recognition (CVPR) , pages 5799–5809, 2021

    2021

  22. [30]

    Nvs-gan: Benefit of generative ad- versarial network on novel view synthesis

    HS Shrisha and V Anupama. Nvs-gan: Benefit of generative ad- versarial network on novel view synthesis. International Journal of Intelligent Networks , 5:184–195, 2024

    2024

  23. [31]

    View independent gen- erative adversarial network for novel view synthesis

    Xiaogang Xu, Ying-Cong Chen, and Jiaya Jia. View independent gen- erative adversarial network for novel view synthesis. In International Conference on Computer Vision (ICCV) , pages 7791–7800, 2019

    2019

  24. [32]

    Rvt: Robotic view transformer for 3d object manipulation

    Ankit Goyal, Jie Xu, Yijie Guo, Valts Blukis, Yu-Wei Chao, and Dieter Fox. Rvt: Robotic view transformer for 3d object manipulation. In Conference on Robot Learning (CoRL) , pages 694–710, 2023

    2023

  25. [33]

    Relative pose estimation with a single affine correspondence

    Banglei Guan, Ji Zhao, Zhang Li, Fang Sun, and Friedrich Fraundorfer. Relative pose estimation with a single affine correspondence. IEEE Transactions on Cybernetics , 52(10):10111–10122, 2022

    2022

  26. [34]

    Minimal solvers for relative pose estimation of multi-camera systems using affine correspondences

    Banglei Guan, Ji Zhao, Daniel Barath, and Friedrich Fraundorfer. Minimal solvers for relative pose estimation of multi-camera systems using affine correspondences. International Journal of Computer Vision (IJCV), 131(1):324–345, 2023

    2023

  27. [35]

    Disentangling factors of variation with cycle-consistent variational auto-encoders

    Ananya Harsh Jha, Saket Anand, Maneesh Singh, and VS Rao Veeravasarapu. Disentangling factors of variation with cycle-consistent variational auto-encoders. In Proceedings of the European Conference on Computer Vision (ECCV) , pages 805–820, 2018

    2018

  28. [36]

    Calvin: A benchmark for language-conditioned policy learning for long-horizon robot manipulation tasks

    Oier Mees, Lukas Hermann, Erick Rosete-Beas, and Wolfram Burgard. Calvin: A benchmark for language-conditioned policy learning for long-horizon robot manipulation tasks. IEEE Robotics and Automation Letters, 7(3):7327–7334, 2022

    2022

  29. [37]

    Minilm: Deep self-attention distillation for task-agnostic com- pression of pre-trained transformers

    Wenhui Wang, Furu Wei, Li Dong, Hangbo Bao, Nan Yang, and Ming Zhou. Minilm: Deep self-attention distillation for task-agnostic com- pression of pre-trained transformers. Advances in Neural Information Processing Systems (NIPS) , 33:5776–5788, 2020

    2020

  30. [38]

    Pseudo-keypoint rkhs learn- ing for self-supervised 6dof pose estimation

    Yangzheng Wu and Michael Greenspan. Pseudo-keypoint rkhs learn- ing for self-supervised 6dof pose estimation. In European Conference on Computer Vision (ECCV) , pages 37–56. Springer, 2024. VI. APPENDIX NOTATIONS AND DEFINITIONS A. Inputs, Outputs, Features, and Weights • a: A...

    2024

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.