Pith. sign in

REVIEW 4 major objections 8 minor 90 references

Deform360 shows 3D particle models with physical priors beat 2D video models on scarce deformable data, while large video models win on scale and zero-shot object generalization.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · grok-4.5

2026-07-11 07:02 UTC pith:ZKJVEFJY

load-bearing objection Big real-world deformable dataset with a clean 2D-vs-3D bake-off; the low-data 3D win is useful but partly inherits the same rigidity/contact priors baked into the particle GT. the 4 major comments →

arxiv 2607.05390 v1 pith:ZKJVEFJY submitted 2026-07-06 cs.RO cs.CV

Deform360: A Massive Multi-view Visuotactile Dataset for Deformable World Models

Hongyu Li , Wanjia Fu , Xiaoyan Cong , Zekun Li , Binghao Huang , Hanxiao Jiang , Xintong He , Yiqing Liang
show 6 more authors
Rao Fu Tao Lu Srinath Sridhar Kevin A. Smith George Konidaris Yunzhu Li
This is my paper
classification cs.RO cs.CV
keywords world modelsdeformable object manipulationvisuotactile perceptionparticle dynamicsaction-conditioned video modelsmulti-view trackingrobot planning
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

Predicting how soft everyday objects move and deform under robot contact is hard because their state spaces are huge and contacts are often hidden. Deform360 supplies the missing real-world fuel: 198 diverse objects, nearly 2,000 interactions, 41 surround cameras and dual tactile grippers, plus dense particle trajectories extracted by a markerless visuotactile pipeline. With this benchmark the authors systematically pit 2D action-conditioned video models against 3D particle models. Physics-informed 3D models dominate when data is limited; large pretrained video models overtake them once the full object diversity is available and also produce higher-fidelity images. The same data further supports contact prediction from vision alone and zero-shot robot planning on new hardware. A reader who wants robots that can fold clothes, squeeze packages or handle cables therefore gains both a concrete ranking of current approaches and a public testbed for the next generation of deformable world models.

Core claim

On the Deform360 benchmark, physics-based 3D particle models outperform learning-based 3D and 2D video models in low-data per-episode and multi-episode regimes thanks to structural priors, while a large pretrained 2D video model achieves superior visual metrics and stronger zero-shot multi-object generalization once trained on the dataset’s full diversity.

What carries the argument

The markerless visuotactile 3D tracking pipeline: per-frame 3D Gaussian Splatting supplies geometry; multi-view 2D tracks are lifted into 3D and refined by a physics-informed objective that enforces temporal coherence, local rigidity, spatial smoothness and soft tactile no-slip constraints, yielding the dense particle trajectories used as ground truth.

Load-bearing premise

The particle ground truth treats tactile pressure as a soft no-slip regularizer and assumes micro-slip was avoided during collection; if slip or plastic flow is common, every model ranking rests on systematically biased trajectories.

What would settle it

Collect a controlled subset of the same objects while deliberately introducing and independently measuring slip (high-speed cameras or multi-axis force sensors), retrain the identical 2D and 3D models, and check whether the reported ranking of Chamfer/track errors reverses or the absolute gaps shrink below the published margins.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • Future deformable world models should fuse explicit 3D structural priors with large-scale visual pretraining to obtain both low-data accuracy and zero-shot generalization.
  • Models trained only on Deform360 can already be dropped into model-predictive control on a different robot and lab for simple deformable planning tasks.
  • Contact events can be predicted from multi-view vision alone at roughly 89 % accuracy, showing that visuotactile coupling is learnable.
  • The public 198-object, 23-million-frame suite becomes a standard benchmark for measuring progress in object-centric deformable dynamics.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Hybrid architectures that distill appearance priors from video models into particle or mesh dynamics could close the remaining fidelity–physics gap without requiring ever-larger pure-3D datasets.
  • Adding shear-force tactile sensing would remove the no-slip assumption and enable models that deliberately exploit or compensate for micro-slip.
  • Surround-view capture may be necessary infrastructure for any future foundation model of deformable objects, because single-view self-occlusion systematically hides the local deformations that matter most for contact.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 8 minor

Summary. The manuscript introduces Deform360, a large-scale real-world multi-view visuotactile dataset for deformable object dynamics: 198 daily-life objects, 1,980 interaction sequences, ~23.3M frames / 215.7 hours from 41 calibrated cameras plus bimanual tactile UMI grippers. A markerless pipeline (per-frame 3DGS geometry, multi-view CoTracker3 lifting, and a physics-informed tracking objective with ARAP, Laplacian, and tactile no-slip terms) produces dense particle trajectories. Using these annotations the authors (i) train a visual-to-contact predictor, (ii) systematically benchmark 3D particle models (PhysTwin, ParticleFormer, PGND) against an action-conditioned 2D video model (Cosmos-Predict2.5) under per-episode, multi-episode, and multi-object generalization, and (iii) show a preliminary zero-shot MPC planning demo on a different robot. The reported insight is a data-regime trade-off: structural 3D priors win in low data; large pretrained video models scale better for zero-shot object generalization and visual fidelity.

Significance. If the resource and evaluation hold, this is a high-value contribution for deformable manipulation and world modeling. Existing real datasets are far smaller, less multi-view, or lack tactile and dense markerless 3D motion (Table 1). The combination of 360° capture, synchronized tactile, open taxonomy across 1D/2D/3D deformables, and a reproducible multi-level generalization protocol fills a clear gap. Strengths include: explicit reconstruction metrics (Table 2), a quantitative visuotactile tracking ablation (5× Chamfer improvement), contact-prediction results, three carefully separated generalization settings, and a cross-lab robot transfer demo. Even if some benchmarking claims need tighter framing, the dataset and annotation pipeline alone are likely to become a standard benchmark for object-centric deformable world models.

major comments (4)
  1. Sec. 4.3, Eqs. (2)–(5) and Tables 3–4: The particle trajectories used as ground truth for all 3D world-model comparisons are obtained by minimizing L_track, which includes an ARAP local-rigidity term (Eq. 2) and a soft tactile no-slip regularizer (Eq. 4; normal pressure only; slip intentionally avoided). These are closely related to the structural priors embedded in PhysTwin’s spring-mass simulator and exploitable by particle models. Superior CD/track-error for PhysTwin (Table 3) and ParticleFormer (Table 4) may therefore partly reflect fidelity to annotation assumptions rather than to true material dynamics—especially for plastic objects or residual micro-slip, which the Limitations section itself flags. Image metrics on LBS-rendered particles do not fully break this dependence. Please (a) discuss this prior-alignment risk explicitly in the main text, (b) report sensitivity of Tables 3–
  2. Tables 3–5: All quantitative world-model results are reported as single point estimates with no standard deviations, confidence intervals, or statistical tests across episodes/objects. Given the diversity of materials and the small number of methods, it is hard to judge whether the reported ranking (e.g., PhysTwin CD 0.014 vs ParticleFormer 0.044 in Table 3 future prediction; Cosmos vs ParticleFormer image metrics in Tables 4–5) is robust. Please add per-category or bootstrap variability, or at least episode-level spreads, for the main metrics that support the structural-priors vs scalability trade-off.
  3. Sec. 5.3 multi-object setting and Table 5 / Fig. 6: Cosmos is compared only on rendered image metrics (PSNR/SSIM/LPIPS), while 3D models also report CD/track error; Cosmos has no particle state. The paper notes a common failure mode—action misalignment over long horizons—yet still credits Cosmos with better zero-shot generalization. Please quantify action-following error (e.g., gripper pose / contact consistency vs commanded actions) for Cosmos under multi-object, and clarify what “better generalization” means when dynamics can look plausible while actions are not followed. Without this, the central trade-off claim between paradigms is only partially supported in the zero-shot regime.
  4. Sec. 5.4 and Fig. 7: Real-world applicability is demonstrated only with PhysTwin in a qualitative MPC setup on a different robot; Cosmos is not deployed, and no success rates, goal-distance curves, or comparison to open-loop / model-free baselines are given. The abstract and contributions list this as evidence of real-world utility. Either expand to quantitative multi-object planning results (success rate, final CD to goal, horizon length) or clearly demote the claim to a qualitative feasibility illustration so it does not over-support the dataset’s “applicability” contribution.
minor comments (8)
  1. Abstract and §1: wording is occasionally ungrammatical (“learning the dynamics overthe2Dpixelspace”, “Asystematic understanding”). A careful copy-edit pass is needed throughout.
  2. Table 1 vs appendix Table 6: the two comparison tables largely duplicate each other with slightly different columns; consolidate or make the appendix table strictly additive (e.g., full action lists only).
  3. Eq. (1): notation mixes Pn,t for single-view lifts and later global particles; define the fusion step (RANSAC velocity field) more formally before Eq. (5).
  4. §5.1: Chamfer units are given as m² (2.71×10⁻⁵ m²); confirm whether this is squared distance and whether particles are in meters; a linear (mm) scale would be more interpretable for robotics readers.
  5. §5.2 contact prediction: report the train/test split (objects vs episodes) and whether views are held out; 88.67% accuracy is hard to interpret without class balance beyond the random baseline.
  6. Fig. 4 caption and main text: “RGB ground truth” for warped point clouds is ambiguous—clarify whether this is multi-view reconstructed geometry or a single reference view.
  7. Appendix B.2: ParticleFormer is a reimplementation; briefly state any hyperparameter deviations from the original paper so others can reproduce the Table 3–5 numbers.
  8. Project website is cited; for archival review, state what will be released (raw multi-view video, tactile, calibrations, particle tracks, training splits, and code for the tracking objective).

Circularity Check

0 steps flagged

Empirical dataset+benchmark paper with no derivation that reduces by construction; minor shared structural priors between annotation losses and 3D baselines are a possible confound, not circularity.

full rationale

Deform360 is a data-collection and empirical benchmarking paper, not a first-principles derivation. The markerless pipeline (Sec. 4.3, Eqs. 1–5) produces particle trajectories by minimizing a standard multi-term objective (Chamfer shape + ARAP local rigidity + Laplacian + soft tactile no-slip) that is then treated as ground-truth for held-out frame/episode/object evaluation of external world models. No claimed prediction or theoretical result is algebraically forced by those losses or by any fitted constant; CD/track-error and image metrics are computed against held-out data in the usual way. Self-citations to PhysTwin and PGND (overlapping authors) supply evaluation baselines whose performance numbers are newly measured on the independent Deform360 corpus; they are not invoked as uniqueness theorems that forbid alternatives. The reported low-data vs. scalability trade-off is therefore an experimental observation, not a tautology. Shared local-rigidity/contact assumptions between the annotation objective and the physics-based 3D models constitute a possible evaluation bias (correctness risk flagged in Limitations), but do not meet any of the six circularity patterns: there is no self-definitional loop, no fitted parameter renamed as prediction, and no load-bearing self-citation chain that collapses the central claim. Score 1 only for the minor, non-load-bearing self-citation of the baselines themselves.

Axiom & Free-Parameter Ledger

3 free parameters · 4 axioms · 1 invented entities

The central claims rest on the fidelity of the markerless tracking pipeline and on standard assumptions of the evaluated world models. Free parameters are the loss weights and a few algorithmic choices; axioms are ordinary computer-vision and physics modeling assumptions plus the no-slip tactile regularizer; no exotic new physical entities are invented.

free parameters (3)
  • λ_local, λ_lap, λ_tactile = 20.0 / 0.1 / 1.0
    Hand-chosen weights (20.0, 0.1, 1.0) in the tracking objective L_track that balance rigidity, smoothness and tactile consistency; they directly affect the particle trajectories used as ground truth.
  • tactile contact radius r
    Radius used to select particles influenced by activated taxels; not given a numeric value in the main text but controls which particles receive the no-slip term.
  • CoTracker3 clip length / stride = 15 / 5
    15-frame clips with stride 5 chosen to handle occlusions; affects track density and identity.
axioms (4)
  • domain assumption Local rigidity (ARAP) and Laplacian smoothness are reasonable soft constraints for the materials in the dataset.
    Invoked in Eqs. 2–3 and Limitations; plastic materials may violate them.
  • ad hoc to paper Tactile sensors provide a usable no-slip regularizer even though they measure only normal pressure and slip is intentionally avoided.
    Sec. 4.3; central to the claimed tracking improvement under gripper occlusion.
  • domain assumption Per-frame 3DGS + multi-view lifting yields geometry accurate enough to serve as ground-truth particles for dynamics evaluation.
    Sec. 4.2–4.3; supported by held-out PSNR but still an approximation.
  • domain assumption Standard image and point-cloud metrics (PSNR, SSIM, LPIPS, Chamfer, track MSE) are sufficient proxies for world-model quality.
    Used throughout Sec. 5.
invented entities (1)
  • Deform360 markerless visuotactile particle trajectories no independent evidence
    purpose: Serve as dense 3D ground truth for training and evaluating deformable world models.
    Produced by the paper’s own pipeline; independent evidence is limited to the reported reconstruction metrics and qualitative warped-point-cloud figures.

pith-pipeline@v1.1.0-grok45 · 29687 in / 3081 out tokens · 26220 ms · 2026-07-11T07:02:35.836753+00:00 · methodology

0 comments
read the original abstract

Predicting object dynamics (i.e., world modeling) is a fundamental challenge for robotic manipulation, and modeling deformable objects presents a particularly difficult case due to their high-dimensional state spaces and complex material properties. While current world models approach this through two distinct paradigms: learning the dynamics over the 2D pixel space or more explicit 3D geometric space. A systematic understanding of their relative strengths and limitations remains elusive due to the lack of diverse, large-scale real-world data. To address this, we present Deform360, a large-scale visuotactile dataset featuring 198 daily-life objects, 1,980 interaction sequences, and over 215 hours of observations from 41 surround-view cameras and bimanual tactile grippers to capture both global motion and contact-induced local deformations. Leveraging a novel markerless visuotactile 3D tracking pipeline to extract dense geometry and motion, we systematically evaluate current state-of-the-art world models, comparing 2D video models against 3D particle models. Finally, we provide a preliminary demonstration indicating the real-world applicability of our dataset by performing robot planning tasks on deformable objects. Our analysis reveals key insights into the trade-offs between structural priors and scalability, providing a solid benchmark for future research in generalizable deformable object-centric world modeling. Project website: https://deform360.lhy.xyz

Figures

Figures reproduced from arXiv: 2607.05390 by Binghao Huang, George Konidaris, Hanxiao Jiang, Hongyu Li, Kevin A. Smith, Rao Fu, Srinath Sridhar, Tao Lu, Wanjia Fu, Xiaoyan Cong, Xintong He, Yiqing Liang, Yunzhu Li, Zekun Li.

Figure 1
Figure 1. Figure 1: Overview of Deform360. We collect a massive multi-view visuotactile dataset with 198 deformable objects (a subset is shown), supporting 2D and 3D world models, contact detection, and real-world robot planning tasks. objects remains a significant challenge [14, 19, 60]. Deformable bodies possess theoretically infinite degrees of freedom, and contact-induced local deformations are frequently occluded by end-… view at source ↗
Figure 2
Figure 2. Figure 2: Overview of our annotation pipeline. We convert multi-view video and tactile streams into dense annotations for deformable object dynamics. We first recon￾struct high-fidelity dynamic geometry using per-frame 3D Gaussian Splatting. Then, we perform markerless 2D tracking, lift the tracks into 3D, and optimize for temporal and multi-view consistency and physical plausibility. folding, and twisting. For each… view at source ↗
Figure 3
Figure 3. Figure 3: Visualization of our particle tracking system across different object categories. 4.1 Visuotactile Preprocessing The pipeline begins with precise multi-view calibration using ArUco grids, en￾suring metric-scale 3D reconstruction. Since the raw RGB videos {Vn} are cap￾tured with lens distortion, we first undistort all camera streams and refer to the undistorted videos as {Vn} for simplicity throughout the p… view at source ↗
Figure 4
Figure 4. Figure 4: Benefit of Visuotactile Perception. The three rows compare the warped point clouds, obtained by temporally integrating per-frame tracking results: our full vi￾suotactile approach, tracking without tactile optimization, and the RGB ground truth. Our method yields superior tracking accuracy under occlusions caused by the gripper, confirming the advantage of integrating tactile feedback into the tracking pipe… view at source ↗
Figure 5
Figure 5. Figure 5: A visualization of the prediction performance from different camera views. 3D World Models. We train learning-based 3D world models, such as ParticleFormer [27] and PGND [86], and optimize the dif￾ferentiable simulation-based PhysT￾win [30] using the dense particle tra￾jectories obtained from our annota￾tion pipeline. We measure the multi￾step prediction error using Chamfer distance (CD) and mean squared e… view at source ↗
Figure 6
Figure 6. Figure 6: Visualization of Multi-Object Generalization [PITH_FULL_IMAGE:figures/full_fig_p013_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Real-world Robot Planning with Model Predictive Control [PITH_FULL_IMAGE:figures/full_fig_p014_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Visualization of dynamic reconstruction results from novel viewpoints and dif￾ferent temporal steps. common representation for deformable objects is the spring-mass system [30,88], where particles are connected by elastic springs. The force Fi on particle i is calculated as the sum of internal spring forces, damping, and external forces: \mathbf {F}_i = \sum _{j \in \mathcal {N}(i)} k_{ij} (\|\mathbf {p}_j… view at source ↗
Figure 9
Figure 9. Figure 9: Visualization of Multi-Episode Generalization. We show the predicted future frames for the glove-cloth (top) and sack-cloth (bottom) objects, comparing different world models under the episode generalization setting. D Dataset Object Taxonomy and Diversity The Deform360 dataset is designed to capture the vast complexity of real-world deformable object dynamics. We categorize the 198 daily-life objects into… view at source ↗
Figure 10
Figure 10. Figure 10: Visualization of 1D Deformables. Object names are marked at the lower right corner of each image [PITH_FULL_IMAGE:figures/full_fig_p032_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Visualization of 2D Deformables. Object names are marked at the lower right corner of each image [PITH_FULL_IMAGE:figures/full_fig_p033_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Visualization of 2D Deformables (continued) [PITH_FULL_IMAGE:figures/full_fig_p034_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: Visualization of 2D Deformables (continued) [PITH_FULL_IMAGE:figures/full_fig_p035_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: Visualization of 3D Deformables. Object names are marked at the lower right corner of each image [PITH_FULL_IMAGE:figures/full_fig_p036_14.png] view at source ↗
Figure 15
Figure 15. Figure 15: Visualization of 3D Deformables (continued) [PITH_FULL_IMAGE:figures/full_fig_p037_15.png] view at source ↗
Figure 16
Figure 16. Figure 16: Visualization of 3D Deformables (continued) [PITH_FULL_IMAGE:figures/full_fig_p038_16.png] view at source ↗
Figure 17
Figure 17. Figure 17: Visualization of Multi-view Deformables. Each image represents a dif￾ferent view of the same object [PITH_FULL_IMAGE:figures/full_fig_p039_17.png] view at source ↗
Figure 18
Figure 18. Figure 18: Visualization of Multi-view Deformables. Each image represents a dif￾ferent view of the same object [PITH_FULL_IMAGE:figures/full_fig_p040_18.png] view at source ↗
Figure 19
Figure 19. Figure 19: Visualization of Multi-view Deformables. Each image represents a dif￾ferent view of the same object [PITH_FULL_IMAGE:figures/full_fig_p041_19.png] view at source ↗

Review history (2 revisions) →

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

90 extracted references · 7 canonical work pages · 1 internal anchor

  1. [1]

    48550/arXiv.2504.03597

    Abou-Chakra, J., Sun, L., Rana, K., May, B., Schmeckpeper, K., Minniti, M.V., Herlant, L.: Real-is-Sim: Bridging the Sim-to-Real Gap with a Dynamic Digital Twin for Real-World Robot Policy Evaluation (Apr 2025).https://doi.org/10. 48550/arXiv.2504.03597

    Pith/arXiv arXiv 2025

  2. [2]

    48550/arXiv.2503.06669

    AgiBot-World-Contributors, Bu, Q., Cai, J., Chen, L., Cui, X., Ding, Y., Feng, S., Gao, S., He, X., Hu, X., Huang, X., Jiang, S., Jiang, Y., Jing, C., Li, H., Li, J., Liu, C., Liu, Y., Lu, Y., Luo, J., Luo, P., Mu, Y., Niu, Y., Pan, Y., Pang, J., Qiao, Y., Ren, G., Ruan, C., Shan, J., Shen, Y., Shi, C., Shi, M., Shi, M., Sima, C., Song, J., Wang, H., Wang...

    Pith/arXiv arXiv 2025

  3. [3]

    Science Robotics10(106), eadt1497 (Sep 2025).https://doi.org/10.1126/ scirobotics.adt1497

    Ai, B., Tian, S., Shi, H., Wang, Y., Pfaff, T., Tan, C., Christensen, H.I., Su, H., Wu, J., Li, Y.: A review of learning-based dynamics models for robotic manipula- tion. Science Robotics10(106), eadt1497 (Sep 2025).https://doi.org/10.1126/ scirobotics.adt1497

    2025

  4. [4]

    In: The Thirteenth International Conference on Learning Representa- tions (Oct 2024)

    Bansal, H., Lin, Z., Xie, T., Zong, Z., Yarom, M., Bitton, Y., Jiang, C., Sun, Y., Chang, K.W., Grover, A.: VideoPhy: Evaluating Physical Commonsense for Video Generation. In: The Thirteenth International Conference on Learning Representa- tions (Oct 2024)

    2024

  5. [5]

    In: Thirty-Fifth Conference on Neural Information Processing Systems Datasets and Benchmarks Track (Round 1) (Jun 2021)

    Bear, D., Wang, E., Mrowca, D., Binder, F.J., Tung, H.Y., Pramod, R.T., Hold- away, C., Tao, S., Smith, K.A., Sun, F.Y., Fei-Fei, L., Kanwisher, N., Tenenbaum, J.B., Yamins, D.L., Fan, J.E.: Physion: Evaluating Physical Prediction from Vi- sion in Humans and Machines. In: Thirty-Fifth Conference on Neural Information Processing Systems Datasets and Benchm...

    2021

  6. [6]

    IEEE Robotics and Automation Letters9(3), 2981–2988 (Mar 2024).https://doi.org/10.1109/LRA

    Blanco-Mulero, D., Barbany, O., Alcan, G., Colomé, A., Torras, C., Kyrki, V.: Benchmarking the Sim-to-Real Gap in Cloth Manipulation. IEEE Robotics and Automation Letters9(3), 2981–2988 (Mar 2024).https://doi.org/10.1109/LRA. 2024.3360814

    doi:10.1109/lra 2024

  7. [7]

    https://doi.org/10.48550/arXiv.2311.15127

    Blattmann, A., Dockhorn, T., Kulal, S., Mendelevitch, D., Kilian, M., Lorenz, D., Levi, Y., English, Z., Voleti, V., Letts, A., Jampani, V., Rombach, R.: Stable Video Diffusion: Scaling Latent Video Diffusion Models to Large Datasets (Nov 2023). https://doi.org/10.48550/arXiv.2311.15127

    Pith/arXiv arXiv doi:10.48550/arxiv.2311.15127 2023

  8. [8]

    Carion, N., Gustafson, L., Hu, Y.T., Debnath, S., Hu, R., Suris, D., Ryali, C., Alwala, K.V., Khedr, H., Huang, A., Lei, J., Ma, T., Guo, B., Kalla, A., Marks, M., Greer, J., Wang, M., Sun, P., Rädle, R., Afouras, T., Mavroudi, E., Xu, K., Wu, T.H., Zhou, Y., Momeni, L., Hazra, R., Ding, S., Vaze, S., Porcher, F., Li, F., Li, S., Kamath, A., Cheng, H.K., ...

    Pith/arXiv arXiv doi:10.48550/arxiv.2511.16719 2025

  9. [9]

    In: 8th Annual Conference on Robot Learning (Sep 2024)

    Chen, Y., Zhang, Y., Brei, Z., Zhang, T., Chen, Y., Wu, J., Vasudevan, R.: Dif- ferentiable Discrete Elastic Rods for Real-Time Modeling of Deformable Linear Objects. In: 8th Annual Conference on Robot Learning (Sep 2024)

    2024

  10. [10]

    Chen, Y., Hu, Y., Sun, L., Kusnur, T., Herlant, L., Jiang, C.: EMPM: Embodied MPM for Modeling and Simulation of Deformable Objects (Jan 2026).https: //doi.org/10.48550/arXiv.2601.17251 Deform360 17

    work page doi:10.48550/arxiv.2601.17251 2026

  11. [11]

    Chi, C., Xu, Z., Pan, C., Cousineau, E., Burchfiel, B., Feng, S., Tedrake, R., Song, S.: Universal Manipulation Interface: In-The-Wild Robot Teaching Without In- The-Wild Robots (Feb 2024).https://doi.org/10.48550/arXiv.2402.10329

    Pith/arXiv arXiv doi:10.48550/arxiv.2402.10329 2024

  12. [12]

    The International Journal of Robotics Research44(9), 1431–1442 (Aug 2025).https://doi.org/10.1177/02783649251317617

    Coltraro, F., Borràs, J., Alberich-Carramiñana, M., Torras, C.: Tracking cloth de- formation: A novel dataset for closing the sim-to-real gap for robotic cloth manip- ulation learning. The International Journal of Robotics Research44(9), 1431–1442 (Aug 2025).https://doi.org/10.1177/02783649251317617

    work page doi:10.1177/02783649251317617 2025

  13. [13]

    In: Proceedings of the IEEE/CVF Interna- tional Conference on Computer Vision

    Cong, X., Xing, A., Pokhariya, C., Fu, R., Sridhar, S.: Dytact: Capturing dynamic contacts in hand-object manipulation. In: Proceedings of the IEEE/CVF Interna- tional Conference on Computer Vision. pp. 7191–7203 (2025)

    2025

  14. [14]

    arXiv preprint arXiv:2406.10167 (2024)

    Cong, X., Yang, H., Chen, L., Zhang, K., Yi, L., Bajaj, C., Huang, Q.: 4drecons: 4d neural implicit deformable objects reconstruction from a single rgb-d camera with geometrical and topological regularizations. arXiv preprint arXiv:2406.10167 (2024)

    Pith/arXiv arXiv 2024

  15. [15]

    arXiv preprint arXiv:2512.16906 (2025)

    Cong, X., Yang, H., Wang, A., Wang, Y., Yang, Y., Zhang, C., Ma, C.: Viva: Vlm- guided instruction-based video editing with reward optimization. arXiv preprint arXiv:2512.16906 (2025)

    arXiv 2025

  16. [16]

    arXiv preprint arXiv:2601.23286 (2026)

    Du, H., Ye, J., Cong, X., Li, R., Ni, J., Agarwal, A., Zhou, Z., Li, Z., Balestriero, R., Wang, Y.: Videogpa: Distilling geometry priors for 3d-consistent video generation. arXiv preprint arXiv:2601.23286 (2026)

    Pith/arXiv arXiv 2026

  17. [17]

    In: Advances in Neural Information Processing Systems (Dec 2023)

    Du, Y., Yang, S., Dai, B., Dai, H., Nachum, O., Tenenbaum, J., Schuurmans, D., Abbeel, P.: Learning Universal Policies via Text-Guided Video Generation. In: Advances in Neural Information Processing Systems (Dec 2023)

    2023

  18. [18]

    In: The Twelfth International Conference on Learning Repre- sentations (Oct 2023)

    Du, Y., Yang, S., Florence, P., Xia, F., Wahid, A., Ichter, B., Sermanet, P., Yu, T., Abbeel, P., Tenenbaum, J.B., Kaelbling, L.P., Zeng, A., Tompson, J.: Video Language Planning. In: The Twelfth International Conference on Learning Repre- sentations (Oct 2023)

    2023

  19. [19]

    https://doi.org/10.48550/arXiv.2312.00583

    Duisterhof, B.P., Mandi, Z., Yao, Y., Liu, J.W., Seidenschwarz, J., Shou, M.Z., Ra- manan, D., Song, S., Birchfield, S., Wen, B., Ichnowski, J.: DeformGS: Scene Flow in Highly Deformable Scenes for Deformable Object Manipulation (Aug 2024). https://doi.org/10.48550/arXiv.2312.00583

    Pith/arXiv arXiv doi:10.48550/arxiv.2312.00583 2024

  20. [20]

    Fu, J., Nan, J., Sun, L., Li, H., Qian, J., Barry, J.L., Kitani, K., Konidaris, G.: NovaPlan: Zero-Shot Long-Horizon Manipulation via Closed-Loop Video Language Planning (Feb 2026).https://doi.org/10.48550/arXiv.2602.20119

    doi:10.48550/arxiv.2602.20119 2026

  21. [21]

    IEEE Robotics and Automation Letters7(3), 5866–5873 (Jul 2022).https://doi.org/ 10.1109/LRA.2022.3158428

    Garcia-Camacho, I., Borràs, J., Calli, B., Norton, A., Alenyà, G.: Household Cloth Object Set: Fostering Benchmarking in Deformable Object Manipulation. IEEE Robotics and Automation Letters7(3), 5866–5873 (Jul 2022).https://doi.org/ 10.1109/LRA.2022.3158428

    work page doi:10.1109/lra.2022.3158428 2022

  22. [22]

    Ha, D., Schmidhuber, J.: World Models (Mar 2018).https://doi.org/10.5281/ zenodo.1207631

    2018

  23. [23]

    Hafner,D.,Lillicrap,T.,Ba,J.,Norouzi,M.:DreamtoControl:LearningBehaviors by Latent Imagination (Mar 2020)

    2020

  24. [24]

    Hafner, D., Pasukonis, J., Ba, J., Lillicrap, T.: Mastering Diverse Domains through World Models (Jan 2023).https://doi.org/10.48550/arXiv.2301.04104

    Pith/arXiv arXiv doi:10.48550/arxiv.2301.04104 2023

  25. [25]

    In: The Twelfth International Conference on Learning Repre- sentations (Oct 2023)

    Hansen, N., Su, H., Wang, X.: TD-MPC2: Scalable, Robust World Models for Continuous Control. In: The Twelfth International Conference on Learning Repre- sentations (Oct 2023)

    2023

  26. [26]

    48550/arXiv.2505.14357 18 H

    Huang, S., Wu, J., Zhou, Q., Miao, S., Long, M.: Vid2World: Crafting Video Dif- fusion Models to Interactive World Models (May 2025).https://doi.org/10. 48550/arXiv.2505.14357 18 H. Li et al

    arXiv 2025

  27. [28]

    Huang, W., Chao, Y.W., Mousavian, A., Liu, M.Y., Fox, D., Mo, K., Fei-Fei, L.: PointWorld: Scaling 3D World Models for In-The-Wild Robotic Manipulation (Jan 2026).https://doi.org/10.48550/arXiv.2601.03782

    doi:10.48550/arxiv.2601.03782 2026

  28. [29]

    Hunyuan3D, T., Zhang, B., Guo, C., Liu, H., Yan, H., Shi, H., Huang, J., Yu, J., Li, K., Linus, Wang, P., Lin, Q., Liu, S., Yang, X., Tang, Y., Zhao, Y., Lai, Z., Liang, Z., Zhao, Z.: Hunyuan3D-Omni: A Unified Framework for Controllable Generation of 3D Assets (Sep 2025).https://doi.org/10.48550/arXiv.2509.21245

    doi:10.48550/arxiv.2509.21245 2025

  29. [30]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision

    Jiang, H., Hsu, H.Y., Zhang, K., Yu, H.N., Wang, S., Li, Y.: PhysTwin: Physics- Informed Reconstruction and Simulation of Deformable Objects from Videos. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 7219–7230 (2025)

    2025

  30. [31]

    In: 2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

    Jin, S., Wang, C., Tomizuka, M.: Robust Deformation Model Approximation for Robotic Cable Manipulation. In: 2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). pp. 6586–6593 (Nov 2019).https: //doi.org/10.1109/IROS40897.2019.8968157

    doi:10.1109/iros40897.2019.8968157 2019

  31. [32]

    Kang, B., Yue, Y., Lu, R., Lin, Z., Zhao, Y., Wang, K., Huang, G., Feng, J.: How Far is Video Generation from World Model: A Physical Law Perspective (Nov 2024).https://doi.org/10.48550/arXiv.2411.02385

    Pith/arXiv arXiv doi:10.48550/arxiv.2411.02385 2024

  32. [33]

    Karaev, N., Makarov, I., Wang, J., Neverova, N., Vedaldi, A., Rupprecht, C.: Co- Tracker3:SimplerandBetterPointTrackingbyPseudo-LabellingRealVideos(Oct 2024).https://doi.org/10.48550/arXiv.2410.11831

    Pith/arXiv arXiv doi:10.48550/arxiv.2410.11831 2024

  33. [34]

    ACM Trans

    Kerbl, B., Kopanas, G., Leimkuehler, T., Drettakis, G.: 3D Gaussian Splatting for Real-Time Radiance Field Rendering. ACM Trans. Graph.42(4), 139:1–139:14 (Jul 2023).https://doi.org/10.1145/3592433

    doi:10.1145/3592433 2023

  34. [35]

    Kong, W., Tian, Q., Zhang, Z., Min, R., Dai, Z., Zhou, J., Xiong, J., Li, X., Wu, B., Zhang, J., Wu, K., Lin, Q., Yuan, J., Long, Y., Wang, A., Wang, A., Li, C., Huang, D., Yang, F., Tan, H., Wang, H., Song, J., Bai, J., Wu, J., Xue, J., Wang, J., Wang, K., Liu, M., Li, P., Li, S., Wang, W., Yu, W., Deng, X., Li, Y., Chen, Y., Cui, Y., Peng, Y., Yu, Z., H...

    Pith/arXiv arXiv doi:10.48550/arxiv.2412.03603 2025

  35. [36]

    Lai, Z., Zhao, Y., Liu, H., Zhao, Z., Lin, Q., Shi, H., Yang, X., Yang, M., Yang, S., Feng, Y., Zhang, S., Huang, X., Luo, D., Yang, F., Yang, F., Wang, L., Liu, S., Tang, Y., Cai, Y., He, Z., Liu, T., Liu, Y., Jiang, J., Linus, Huang, J., Guo, C.: Hunyuan3D 2.5: Towards High-Fidelity 3D Assets Generation with Ultimate Details (Jun 2025).https://doi.org/1...

    Pith/arXiv arXiv doi:10.48550/arxiv.2506.16504 2025

  36. [37]

    https://openreview.net/forum?id=BZ5a1r-kVsf (Jun 2022)

    LeCun, Y.: A Path Towards Autonomous Machine Intelligence. https://openreview.net/forum?id=BZ5a1r-kVsf (Jun 2022)

    2022

  37. [38]

    Li, D., Fang, Y., Chen, Y., Yang, S., Cao, S., Wong, J., Luo, M., Wang, X., Yin, H., Gonzalez, J.E., Stoica, I., Han, S., Lu, Y.: WorldModelBench: Judging Video Generation Models As World Models (Feb 2025).https://doi.org/10.48550/ arXiv.2502.20694

    Pith/arXiv arXiv 2025

  38. [39]

    In: 2025 International Conference on Robotics and Automation (ICRA) (Apr 2025) Deform360 19

    Li, H., Akl, J., Sridhar, S., Brady, T., Padir, T.: ViTa-Zero: Zero-shot Visuotactile Object 6D Pose Estimation. In: 2025 International Conference on Robotics and Automation (ICRA) (Apr 2025) Deform360 19

    2025

  39. [40]

    In: 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

    Li, H., Dikhale, S., Cui, J., Iba, S., Jamali, N.: HyperTaxel: Hyper-Resolution for Taxel-Based Tactile Signals Through Contrastive Learning. In: 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). pp. 7499– 7506 (Oct 2024).https://doi.org/10.1109/IROS58592.2024.10802001

    doi:10.1109/iros58592.2024.10802001 2024

  40. [41]

    In: Robotics: Science and Systems (RSS) XXI (Feb 2025)

    Li, H., Jia, M., Akbulut, T., Xiang, Y., Konidaris, G., Sridhar, S.: V-HOP: Visuo- Haptic 6D Object Pose Tracking. In: Robotics: Science and Systems (RSS) XXI (Feb 2025)

    2025

  41. [42]

    Li, H., Sun, L., Hu, Y., Ta, D., Barry, J., Konidaris, G., Fu, J.: NovaFlow: Zero- Shot Manipulation via Actionable Flow from Generated Videos (Oct 2025).https: //doi.org/10.48550/arXiv.2510.08568

    doi:10.48550/arxiv.2510.08568 2025

  42. [43]

    IEEE Transactions on Pattern Analysis and Machine Intelligence47(9), 7243–7254 (Sep 2025).https://doi.org/10

    Li, X., Guo, Y., Tu, Y., Ji, Y., Liu, Y., Ye, J., Zheng, C.: Textureless Deformable Object Tracking With Invisible Markers. IEEE Transactions on Pattern Analysis and Machine Intelligence47(9), 7243–7254 (Sep 2025).https://doi.org/10. 1109/TPAMI.2024.3463422

    arXiv 2025

  43. [44]

    ACM Trans

    Li, Y., Du, T., Wu, K., Xu, J., Matusik, W.: DiffCloth: Differentiable Cloth Simula- tion with Dry Frictional Contact. ACM Trans. Graph.42(1), 2:1–2:20 (Oct 2022). https://doi.org/10.1145/3527660

    doi:10.1145/3527660 2022

  44. [45]

    Li, Y., Wu, J., Tedrake, R., Tenenbaum, J.B., Torralba, A.: Learning Particle Dynamics for Manipulating Rigid Bodies, Deformable Objects, and Fluids (Apr 2019).https://doi.org/10.48550/arXiv.1810.01566

    Pith/arXiv arXiv doi:10.48550/arxiv.1810.01566 2019

  45. [46]

    In: 2019 In- ternational Conference on Robotics and Automation (ICRA)

    Li, Y., Wu, J., Zhu, J.Y., Tenenbaum, J.B., Torralba, A., Tedrake, R.: Propaga- tion Networks for Model-Based Control Under Partial Observation. In: 2019 In- ternational Conference on Robotics and Automation (ICRA). pp. 1205–1211 (May 2019).https://doi.org/10.1109/ICRA.2019.8793509

    work page doi:10.1109/icra.2019.8793509 2019

  46. [47]

    Liang, L., Bian, L., Xiao, C., Zhang, J., Chen, L., Liu, I., Xiang, F., Huang, Z., Su, H.: Robo360: A 3D Omnispective Multi-Material Robotic Manipulation Dataset (Dec 2023).https://doi.org/10.48550/arXiv.2312.06686

    Pith/arXiv arXiv doi:10.48550/arxiv.2312.06686 2023

  47. [48]

    https://arxiv.org/abs/2307.16897v1 (Jul 2023)

    Lu, C.Y., Zhou, P., Xing, A., Pokhariya, C., Dey, A., Shah, I., Mavidipalli, R., Hu, D., Comport, A., Chen, K., Sridhar, S.: DiVA-360: The Dynamic Visuo- Audio Dataset for Immersive Neural Fields. https://arxiv.org/abs/2307.16897v1 (Jul 2023)

    Pith/arXiv arXiv 2023

  48. [49]

    Luiten, J., Kopanas, G., Leibe, B., Ramanan, D.: Dynamic 3D Gaussians: Tracking by Persistent Dynamic View Synthesis (Aug 2023).https://doi.org/10.48550/ arXiv.2308.09713

    Pith/arXiv arXiv 2023

  49. [50]

    IEEE Access12, 97581–97604 (2024).https: //doi.org/10.1109/ACCESS.2024.3425448

    Newbury, R., Collins, J., He, K., Pan, J., Posner, I., Howard, D., Cosgun, A.: A Review of Differentiable Simulators. IEEE Access12, 97581–97604 (2024).https: //doi.org/10.1109/ACCESS.2024.3425448

    doi:10.1109/access.2024.3425448 2024

  50. [51]

    Li et al

    NVIDIA, Agarwal, N., Ali, A., Bala, M., Balaji, Y., Barker, E., Cai, T., Chat- topadhyay, P., Chen, Y., Cui, Y., Ding, Y., Dworakowski, D., Fan, J., Fenzi, M., Ferroni, F., Fidler, S., Fox, D., Ge, S., Ge, Y., Gu, J., Gururani, S., He, E., Huang, J., Huffman, J., Jannaty, P., Jin, J., Kim, S.W., Klár, G., Lam, G., Lan, S., Leal- Taixe, L., Li, A., Li, Z.,...

    Pith/arXiv arXiv doi:10.48550/arxiv.2501.03575 2025

  51. [52]

    IEEE Robotics and Automation Letters10(10), 10586–10593 (Oct 2025).https://doi.org/10.1109/LRA.2025.3600148

    Obrist, J., Zamora, M., Zheng, H., Hinchet, R., Ozdemir, F., Zarate, J., Katzschmann, R.K., Coros, S.: PokeFlex: A Real-World Dataset of Volumetric Deformable Objects for Robotics. IEEE Robotics and Automation Letters10(10), 10586–10593 (Oct 2025).https://doi.org/10.1109/LRA.2025.3600148

    doi:10.1109/lra.2025.3600148 2025

  52. [53]

    Transactions on Machine Learning Research (Jul 2023)

    Oquab, M., Darcet, T., Moutakanni, T., Vo, H.V., Szafraniec, M., Khalidov, V., Fernandez, P., Haziza, D., Massa, F., El-Nouby, A., Assran, M., Ballas, N., Galuba, W.,Howes,R.,Huang,P.Y.,Li,S.W.,Misra,I.,Rabbat,M.,Sharma,V.,Synnaeve, G., Xu, H., Jegou, H., Mairal, J., Labatut, P., Joulin, A., Bojanowski, P.: DINOv2: Learning Robust Visual Features without ...

    2023

  53. [54]

    https://doi.org/10.48550/arXiv.2507.00990

    Patel, S., Mohan, S., Mai, H., Jain, U., Lazebnik, S., Li, Y.: Robotic Manipula- tion by Imitating Generated Videos Without Physical Demonstrations (Jul 2025). https://doi.org/10.48550/arXiv.2507.00990

    Pith/arXiv arXiv doi:10.48550/arxiv.2507.00990 2025

  54. [55]

    https://doi.org/10.48550/arXiv.2212.09748

    Peebles, W., Xie, S.: Scalable Diffusion Models with Transformers (Mar 2023). https://doi.org/10.48550/arXiv.2212.09748

    Pith/arXiv arXiv doi:10.48550/arxiv.2212.09748 2023

  55. [56]

    In: Proceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Pokhariya, C., Shah, I.N., Xing, A., Li, Z., Chen, K., Sharma, A., Sridhar, S.: MANUS: Markerless Grasp Capture using Articulated 3D Gaussians. In: Proceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 2197–2208 (2024)

    2024

  56. [57]

    In: Pro- ceedings of The 6th Conference on Robot Learning

    Qin, Y., Huang, B., Yin, Z.H., Su, H., Wang, X.: DexPoint: Generalizable Point Cloud Reinforcement Learning for Sim-to-Real Dexterous Manipulation. In: Pro- ceedings of The 6th Conference on Robot Learning. pp. 594–605. PMLR (Mar 2023)

    2023

  57. [58]

    In: Proceedings of the 37th International Conference on Machine Learning

    Sanchez-Gonzalez, A., Godwin, J., Pfaff, T., Ying, R., Leskovec, J., Battaglia, P.: Learning to Simulate Complex Physics with Graph Networks. In: Proceedings of the 37th International Conference on Machine Learning. pp. 8459–8468. PMLR (Nov 2020)

    2020

  58. [59]

    SIGART Bull.2(4), 160–163 (Jul 1991).https://doi.org/10.1145/122344

    Sutton, R.S.: Dyna, an integrated architecture for learning, planning, and react- ing. SIGART Bull.2(4), 160–163 (Jul 1991).https://doi.org/10.1145/122344. 122377

    doi:10.1145/122344 1991

  59. [60]

    The International Journal of Robotics Research41(6), 599– 614 (May 2022).https://doi.org/10.1177/0278364919841431

    Tang, T., Tomizuka, M.: Track deformable objects from point clouds with structure preserved registration. The International Journal of Robotics Research41(6), 599– 614 (May 2022).https://doi.org/10.1177/0278364919841431

    work page doi:10.1177/0278364919841431 2022

  60. [61]

    Team, G.R., Abdolmaleki, A., Abeyruwan, S., Ainslie, J., Alayrac, J.B., Are- nas, M.G., Balakrishna, A., Batchelor, N., Bewley, A., Bingham, J., Bloesch, M., Bousmalis, K., Brakel, P., Brohan, A., Buschmann, T., Byravan, A., Cabi, S., Caluwaerts, K., Casarini, F., Chan, C., Chang, O., Chappellet-Volpini, L., Chen, J.E., Chen, X., Chiang, H.T.L., Choromans...

    2025

  61. [62]

    Team, G.R., Choromanski, K., Devin, C., Du, Y., Dwibedi, D., Gao, R., Jindal, A., Kipf, T., Kirmani, S., Leal, I., Liu, F., Majumdar, A., Marmon, A., Parada, C., Rubanova, Y., Shah, D., Sindhwani, V., Tan, J., Xia, F., Xiao, T., Yang, S., Yu, W., Zhou, A.: Evaluating Gemini Robotics Policies in a Veo World Simulator (Jan 2026).https://doi.org/10.48550/arX...

    doi:10.48550/arxiv.2512.10675 2026

  62. [63]

    Team, M.L., Cai, X., Huang, Q., Kang, Z., Li, H., Liang, S., Ma, L., Ren, S., Wei, X., Xie, R., Zhang, T.: LongCat-Video Technical Report (Oct 2025).https: //doi.org/10.48550/arXiv.2510.22200

    doi:10.48550/arxiv.2510.22200 2025

  63. [64]

    2511.09057

    Team, P.A.N., Xiang, J., Gu, Y., Liu, Z., Feng, Z., Gao, Q., Hu, Y., Huang, B., Liu, G., Yang, Y., Zhou, K., Abrahamyan, D., Ahmad, A., Bannur, G., Chen, J., Chen, K., Deng, M., Han, R., Huang, X., Kang, H., Liu, Z., Ma, E., Ren, H., Shinde, Y., Shingre, R., Tanikella, R., Tao, K., Yang, D., Yu, X., Zeng, C., Zhou, B., Liu, Z., Hu, Z., Xing, E.P.: PAN: A ...

    doi:10.48550/arxiv 2025

  64. [65]

    Team, S.D., Chen, X., Chu, F.J., Gleize, P., Liang, K.J., Sax, A., Tang, H., Wang, W., Guo, M., Hardin, T., Li, X., Lin, A., Liu, J., Ma, Z., Sagar, A., Song, B., Wang, X., Yang, J., Zhang, B., Dollár, P., Gkioxari, G., Feiszli, M., Malik, J.: SAM 3D: 3Dfy Anything in Images (Nov 2025).https://doi.org/10.48550/arXiv.2511. 16624

    doi:10.48550/arxiv.2511 2025

  65. [66]

    Advances in Neural Information Processing Systems36, 67048–67068 (Dec 2023)

    Tung, H.Y., Ding, M., Chen, Z., Bear, D., Gan, C., Tenenbaum, J., Yamins, D., Fan, J., Smith, K.: Physion++: Evaluating Physical Scene Understanding that Requires Online Inference of Different Physical Properties. Advances in Neural Information Processing Systems36, 67048–67068 (Dec 2023)

    2023

  66. [67]

    In: Advances in Neural Information Processing Systems

    Vaswani,A.,Shazeer,N.,Parmar,N.,Uszkoreit,J.,Jones,L.,Gomez,A.N.,Kaiser, L., Polosukhin, I.: Attention is All you Need. In: Advances in Neural Information Processing Systems. vol. 30 (2017)

    2017

  67. [68]

    Wan, T., Wang, A., Ai, B., Wen, B., Mao, C., Xie, C.W., Chen, D., Yu, F., Zhao, H., Yang, J., Zeng, J., Wang, J., Zhang, J., Zhou, J., Wang, J., Chen, J., Zhu, K., Zhao, K., Yan, K., Huang, L., Feng, M., Zhang, N., Li, P., Wu, P., Chu, R., Feng, R., Zhang, S., Sun, S., Fang, T., Wang, T., Gui, T., Weng, T., Shen, T., Lin, W., Wang, W., Wang, W., Zhou, W.,...

    Pith/arXiv arXiv doi:10.48550/arxiv.2503.20314 2025

  68. [69]

    Li et al

    Wang, C., Zhang, Y., Zhang, X., Wu, Z., Zhu, X., Jin, S., Tang, T., Tomizuka, M.: Offline-Online Learning of Deformation Model for Cable Manipulation With 22 H. Li et al. Graph Neural Networks. IEEE Robotics and Automation Letters7(2), 5544–5551 (Apr 2022).https://doi.org/10.1109/LRA.2022.3158376

    work page doi:10.1109/lra.2022.3158376 2022

  69. [70]

    arXiv preprint arXiv:2603.12231 (2026)

    Wang, Y., Bounou, O., Zhou, G., Balestriero, R., Rudner, T.G., LeCun, Y., Ren, M.: Temporal straightening for latent planning. arXiv preprint arXiv:2603.12231 (2026)

    arXiv 2026

  70. [71]

    Wang, Y., Syed, R., Wu, F., Zhang, M., Onol, A., Barreiros, J., Nayyeri, H., Dear, T., Zhang, H., Li, Y.: Interactive World Simulator for Robot Policy Training and Evaluation (Mar 2026).https://doi.org/10.48550/arXiv.2603.08546

    doi:10.48550/arxiv.2603.08546 2026

  71. [72]

    Wen, B., Tremblay, J., Blukis, V., Tyree, S., Muller, T., Evans, A., Fox, D., Kautz, J., Birchfield, S.: BundleSDF: Neural 6-DoF Tracking and 3D Reconstruction of Unknown Objects (Mar 2023).https://doi.org/10.48550/arXiv.2303.14158

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2303.14158 2023

  72. [73]

    Wiedemer, T., Li, Y., Vicol, P., Gu, S.S., Matarese, N., Swersky, K., Kim, B., Jaini, P., Geirhos, R.: Video models are zero-shot learners and reasoners (Sep 2025).https://doi.org/10.48550/arXiv.2509.20328

    Pith/arXiv arXiv doi:10.48550/arxiv.2509.20328 2025

  73. [74]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Wu, G., Yi, T., Fang, J., Xie, L., Zhang, X., Wei, W., Liu, W., Tian, Q., Wang, X.: 4D Gaussian Splatting for Real-Time Dynamic Scene Rendering. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 20310–20320 (2024)

    2024

  74. [75]

    Wu, P., Escontrela, A., Hafner, D., Goldberg, K., Abbeel, P.: DayDreamer: World Models for Physical Robot Learning (Jun 2022).https://doi.org/10.48550/ arXiv.2206.14176

    Pith/arXiv arXiv 2022

  75. [76]

    In: Proceedings of the IEEE/CVF Confer- ence on Computer Vision and Pattern Recognition

    Wu, P., Zhu, K., Liu, Y., Zhao, L., Zhai, W., Cao, Y., Zha, Z.J.: Improved Video VAE for Latent Video Diffusion Model. In: Proceedings of the IEEE/CVF Confer- ence on Computer Vision and Pattern Recognition. pp. 18124–18133 (2025)

    2025

  76. [77]

    In: The Eleventh International Conference on Learning Representations (Sep 2022)

    Xian, Z., Zhu, B., Xu, Z., Tung, H.Y., Torralba, A., Fragkiadaki, K., Gan, C.: FluidLab: A Differentiable Environment for Benchmarking Complex Fluid Manip- ulation. In: The Eleventh International Conference on Learning Representations (Sep 2022)

    2022

  77. [78]

    https://doi.org/10.48550/arXiv.2412.01506

    Xiang, J., Lv, Z., Xu, S., Deng, Y., Wang, R., Zhang, B., Chen, D., Tong, X., Yang, J.: Structured 3D Latents for Scalable and Versatile 3D Generation (Dec 2024). https://doi.org/10.48550/arXiv.2412.01506

    Pith/arXiv arXiv doi:10.48550/arxiv.2412.01506 2024

  78. [79]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Xie, T., Zong, Z., Qiu, Y., Li, X., Feng, Y., Yang, Y., Jiang, C.: PhysGaussian: Physics-Integrated 3D Gaussians for Generative Dynamics. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 4389– 4398 (2024)

    2024

  79. [80]

    Graphical Models127, 101178 (May 2023).https://doi.org/10.1016/j.gmod.2023.101178

    Xie, W., Yu, Z., Zhao, Z., Zuo, B., Wang, Y.: HMDO : Markerless multi-view hand manipulation capture with deformable objects. Graphical Models127, 101178 (May 2023).https://doi.org/10.1016/j.gmod.2023.101178

    doi:10.1016/j.gmod.2023.101178 2023

  80. [81]

    Xie, X., Wen, B., Chang, Y., Rabeti, H., Li, J., Yuan, Y., Pons-Moll, G., Birchfield, S.: CARI4D: Category Agnostic 4D Reconstruction of Human-Object Interaction (Jan 2026).https://doi.org/10.48550/arXiv.2512.11988

    Pith/arXiv arXiv doi:10.48550/arxiv.2512.11988 2026

Showing first 80 references.