Pith. sign in

REVIEW 3 major objections 5 minor 1 cited by

This survey argues that in 4D content modeling, the choice of representation is the primary design decision, and it supplies a task-oriented framework of geometry, motion, and interaction with explicit trade-offs to guide that choice.

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 · deepseek-v4-flash

2026-08-04 08:41 UTC pith:3JBT2HE4

load-bearing objection A useful representation-centric survey whose practical selection table is weaker than its taxonomy. the 3 major comments →

arxiv 2510.19255 v3 pith:3JBT2HE4 submitted 2025-10-22 cs.CV

Advances in 4D Representation: Geometry, Motion, and Interaction

Mingrui Zhao , Sauradip Nag , Kai Wang , Aditya Vora , Guangda Ji , Peter Chun , Ali Mahdavi-Amiri , Hao Zhang This is my paper
classification cs.CV
keywords 4D representationdynamic scene reconstruction4D generationneural radiance fields3D Gaussian splattingmotion modelinginteraction modelingstructured vs unstructured representations
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.

The survey's central claim is that the field of 4D generation and reconstruction is best understood through its representations, not its applications or algorithms. To make this concrete, it organizes 4D representations along three pillars — geometry (structured vs. unstructured), motion (articulation, deformation, tracking, hybrid), and interaction (pose, contact, action/affordance, physics) — and compares them across seven dimensions in a single trade-off table. The message the authors want readers to take away is that representation choice should be driven by the task's computation, application, and data constraints, and that rendering-oriented workhorses like NeRF and 3D Gaussian Splatting are not automatically the right default for editing and interaction tasks, where structured representations shine. A sympathetic reader would care because, if the framing holds, it converts a scattered literature into a decision framework and points to structured representations as the under-explored growth area.

Core claim

On the paper's own terms, the central discovery is a representation-first map of the 4D landscape: a taxonomy built on three pillars — geometry (meshes, point clouds, NeRF, 3DGS, templates, parts, graphs), motion (articulated, deformation, tracking, hybrid), and interaction (pose, contact, action and affordance, physics) — with a structured/unstructured distinction as its backbone. The load-bearing comparison is Table 2, which rates each geometric representation on visual fidelity, scalability, temporal consistency, topology handling, editability, generalization, and efficiency. The paper argues that unstructured representations excel at novel-view synthesis from sparse inputs, while structu

What carries the argument

The central object is the taxonomy itself, together with the structured-vs-unstructured distinction and the seven-dimension comparison in Table 2. The taxonomy splits geometry into unstructured representations (mesh, point cloud, NeRF, 3D Gaussian Splatting), whose primitives carry no functional or semantic meaning, and structured representations (template, part, graph), which impose explicit compositional constraints; motion is then divided into articulated, deformation, tracking, and hybrid classes; interaction is organized as pose, contact, action/affordance, and physics. The taxonomy is doing the argumentative work: it is the device that converts a large body of methods into a small set

Load-bearing premise

The trade-off ratings in Table 2 assume that the survey's selective, representative sample of methods fairly spans each representation class, so the High/Medium/Low assignments would not hold if the chosen works are unrepresentative of their categories.

What would settle it

Run the seven Table 2 dimensions as a quantitative study: take a balanced sample of methods from each of the seven geometry classes, evaluate them on a common set of dynamic scenes with standardized protocols (e.g., fixed compute, identical sparse-input monocular video), and check whether the relative orderings — point clouds highest scalability, templates highest editability, NeRF highest visual fidelity — reproduce. A single class flipping its rank would undercut the paper's central comparative claim.

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

If this is right

  • If the framework is right, a practitioner can use Table 2 to pick a representation by task: point clouds for large-scale sensor capture, templates or parts for category-level editing and animation, NeRF or 3DGS for high-fidelity novel-view synthesis.
  • It implies that structured representations — templates, part-based models, scene graphs — will become a focus of 4D research for editing and interaction workloads, since they score highest on editability and temporal consistency.
  • It supports the push toward hybrid representations that combine the interpretability of structured models with the flexibility of implicit neural fields.
  • It diagnoses the field's dataset bottleneck: existing data lacks the diversity in motion and interaction, and the geometry ground truth, needed to train representation-aware 4D models.
  • It predicts continued migration from per-scene optimization to feed-forward and SDS-free training, which changes which representations are practical to deploy.

Where Pith is reading between the lines

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

  • The taxonomy's trade-off table could be made falsifiable by running a pantheon of representative methods on a shared set of dynamic scenes and measuring the seven dimensions; the paper does not provide that benchmark, so the ratings are testable hypotheses rather than measurements.
  • If representation choice is truly primary, then evaluation metrics should become representation-aware — for example, measuring editability and temporal consistency alongside image fidelity — otherwise cross-method comparisons remain confounded by representational differences.
  • An implicit consequence of the argument is that the current dominance of NeRF and 3DGS in 4D work may be a historical artifact of their success in static 3D, and that part-aware or template-aware extensions of these representations are the most promising route to combine fidelity with editability.
  • The survey's claim about the role of LLMs and video foundation models as data amplifiers implies that the next bottleneck will be 4D evaluation and dataset curation, not generation quality — an area the paper itself flags as underdeveloped.

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

3 major / 5 minor

Summary. The paper surveys 4D generation and reconstruction from a representation-centric perspective, organizing content into geometry (unstructured: mesh, point clouds, NeRF, 3DGS; structured: template, part, graph), motion (articulated, deformation, tracking, hybrid), and interaction (pose, contact, action/affordance, physics). It includes method tables, a seven-dimension qualitative comparison of representations, datasets/benchmarks, and training-strategy sections. The stated goal is to help readers select and customize appropriate 4D representations for their tasks.

Significance. The representation-centric organization is a useful contribution: the structured/unstructured distinction is coherent, the motion taxonomy is sensible, and the formal definitions (LBS, deformation fields, scene flow, SDS) are standard and correctly stated. The coverage of datasets, benchmarks, and metrics is a practical strength that will help newcomers. However, the promised decision framework rests on Table 2, whose qualitative ratings currently conflate representation-level and method-level evidence. This is fixable but requires substantive revision of the comparison methodology.

major comments (3)
  1. [Table 2, §7 (Generalization row)] Generalization is defined as 'Transferability to unseen scenes/objects without per-instance optimization or retraining,' a method-level property, but the row rates representations. Point Cloud is rated High citing DUST3R/VGGT, which are trained feed-forward models; NeRF and 3DGS are rated Medium even though the survey cites feed-forward variants (LRM, 4D-LRM, L4GM) with strong cross-instance generalization; Graph/Part/Template are Very High largely due to category-specific priors. The accompanying prose then discusses deformation-field motion representations and zero-shot tracking, which are not rows in Table 2. Since the abstract and Section 1 promise selection guidance, this conflation is load-bearing. Please split representation-level inductive bias from method-level generalization, or add a method column.
  2. [Table 2 / §1 (evidence and selection criteria)] Table 2 assigns High/Medium/Low ratings across seven dimensions without an evaluation protocol, benchmark, or quantitative support; Section 1 only says the survey takes a selective approach. Some ratings sit uneasily with the paper's own sections: NeRF is Very High for visual fidelity despite Section 2.1.3 noting persistent flickering and unrealistic deformations, and Mesh is Low for efficiency despite native rasterization and skinning. A reader cannot tell whether a rating reflects the representation or the representative methods. Please specify rating criteria or reframe the table as an informal summary with stated caveats.
  3. [Figure 3 / Table 1 / Table 2 (representative works)] The selective-example basis for the global comparison is not documented. Several works used as representative examples are from the authors' own groups (In-2-4D, SINGAPO, ATOP, 4D-PSG, SweepNet). This is not inherently problematic, but absent inclusion criteria it creates a risk that the trade-off conclusions in Table 2 are driven by convenience samples. Please state the selection protocol or explicitly bound the claims to the selected methods.
minor comments (5)
  1. [§7 (opening sentence)] The section opens by saying the comparison is across 'six key metrics,' but seven dimensions are defined and listed in Table 2. Please align the count and definitions.
  2. [Eq. (3)] The left-hand side is written as Δb→f(pb) but the right-hand side applies Φθ to p_f. The argument should be p_b (or the notation clarified) to match the forward canonical-to-observation mapping.
  3. [Table 1] The legend lists motion types ART, DF, TRK, ST, and PF, but the 4D-LRM row uses 'TG.' This appears to be a typo (likely ST).
  4. [Throughout] There are numerous typos and nonstandard reference abbreviations, e.g., 'Gassuain,' 'syncrhonised,' 'disscused,' 'adpot,' and entries such as [Y*23], [L*19], [WW A*25], [YYj*25] where the abbreviated author list obscures the citation. A thorough proofread and reference cleanup is needed.
  5. [Figure 3] The taxonomy diagram is dense, with small venue labels and numbers overlaid on the tree. The font/contrast should be improved for readability.

Circularity Check

0 steps flagged

No circularity: the survey's claims are organizational and qualitative, not derived from its inputs or from the cited works' outputs.

full rationale

This is a survey with an organizing thesis, not a derivation chain. The paper states in §1: 'we take a more selective approach by focusing on representative works to highlight both the desirable properties and ensuing challenges of each 4D representation' — this is an editorial framing, not a claim derived from equations. Table 2's qualitative High/Medium/Low ratings are presented as comparative judgments ('Table 2 outlines the fundamental trade-offs'), not as predictions fitted from data or obtained by substitution; no equation in the paper reduces one rating to another by construction. The 'Generalization' row's conflation of representation-level and method-level evidence is a legitimate correctness/selection-bias concern, but it is not circularity: the ratings are not defined in terms of the outputs they purport to explain. Several works from the authors' own group (e.g., In-2-4D [NCOZMA25], SINGAPO [LIC*25], ATOP [VNZ25], 4D-PSG [YCP*23]) appear as representative examples in Figure 3, Table 1, and the text, but they are not invoked as a uniqueness theorem, an exclusive justification for the taxonomy, or a fitted input disguised as a prediction. Self-citation of representative examples may raise representativeness concerns, but the survey's central claim — that 4D representations can be organized by geometry, motion, and interaction and selected via trade-offs — does not logically depend on the specific outputs of those papers. No circular step meeting the quoted-evidence bar was found.

Axiom & Free-Parameter Ledger

0 free parameters · 3 axioms · 0 invented entities

The survey's conclusions rest on three editorial premises: that a selective set of representative works spans the design space, that structured vs unstructured is a meaningful partition, and that the qualitative trade-off ratings in Table 2 are trustworthy. All three are framing assumptions rather than derived results. No fitted constants or new entities are introduced.

axioms (3)
  • domain assumption The selective set of representative works covers the relevant 4D representation space for the conclusions drawn.
    Section 1: 'instead of offering an exhaustive enumeration of many works, we take a more selective approach by focusing on representative works.' The survey's trade-off conclusions depend on this coverage.
  • ad hoc to paper Representations can be cleanly categorized as structured vs unstructured by whether primitives carry explicit functional, hierarchical, or semantic meaning.
    Section 2 opening: 'We categorize representations into unstructured and structured classes on whether their operational primitives carry explicit functional, hierarchical, or semantic meaning.' This binary split is author-defined and drives the survey's comparison.
  • domain assumption Qualitative ratings in Table 2 (fidelity, scalability, etc.) can be compared across methods without a standardized benchmark.
    Section 7 and Table 2 assign High/Medium/Low ratings to seven representation families. The survey provides no per-method quantitative evidence or evaluation protocol to justify these ratings.

pith-pipeline@v1.3.0-alltime-deepseek · 57654 in / 10415 out tokens · 91328 ms · 2026-08-04T08:41:38.138095+00:00 · methodology

0 comments
read the original abstract

We present a survey on 4D generation and reconstruction, a fast-evolving subfield of computer graphics whose developments have been propelled by recent advances in neural fields, geometric and motion deep learning, as well as 3D generative artificial intelligence (GenAI). While our survey is not the first of its kind, we build our coverage of the domain from a unique and distinctive perspective of 4D representations, to model 3D geometry evolving over time while exhibiting motion and interaction. Specifically, instead of offering an exhaustive enumeration of many works, we take a more selective approach by focusing on representative works to highlight both the desirable properties and ensuing challenges of each representation under different computation, application, and data scenarios. The main take-away message we aim to convey to the readers is on how to select and then customize the appropriate 4D representations for their tasks. Organizationally, we separate the 4D representations based on three key pillars: geometry, motion, and interaction. Our discourse will not only encompass the most popular representations of today, such as neural radiance fields (NeRFs) and 3D Gaussian Splatting (3DGS), but also bring attention to relatively under-explored representations in the 4D context, such as structured models and long-range motions. Throughout our survey, we will reprise the role of large language models (LLMs) and video foundational models (VFMs) in a variety of 4D applications, while steering our discussion towards their current limitations and how they can be addressed. We also provide a dedicated coverage on what 4D datasets are currently available, as well as what is lacking, in driving the subfield forward. Project page:https://mingrui-zhao.github.io/4DRep-GMI/

Figures

Figures reproduced from arXiv: 2510.19255 by Aditya Vora, Ali Mahdavi-Amiri, Guangda Ji, Hao Zhang, Kai Wang, Mingrui Zhao, Peter Chun, Sauradip Nag.

Figure 1
Figure 1. Figure 1: Representative applications for three key pillars of 4D Representation: (a) (Geometry) mesh animation [CZTW25], (b) (Geom￾etry) dynamic point map reconstruction [WZZ∗ 25], (c) (Motion) 4D inbetweening [NCOZMA25],(d) (Motion) 4D semantic segmentation [MRH22], (e) Human-scene interaction [LYLW24],(f) Human-object interaction [LWL23]. Abstract We present a survey on 4D generation and reconstruction, a fast-ev… view at source ↗
Figure 2
Figure 2. Figure 2: 4D content created with different geometric representations. (a) Mesh [WYWB25]; (b) Point cloud [WZZ∗ 25]; (c) Gaussian Splatting [NCOZMA25]; (d) NeRF [VYB∗ 24]; (e) Template-based representation [ZZY∗ 24]; (f) Part-based representation [LTMAS24]; (g) Spatial-Temporal scene graph [YCP∗ 23]. Figures adopted from the original papers. and benchmarking frameworks that enable systematic comparison across repres… view at source ↗
Figure 3
Figure 3. Figure 3: Our taxonomy of 4D representations. We separate them into three pillars: (1) Geometry, including both structured and unstructured representations; (2) motion, including articulation, deformation and tracking based representations; (3) Interaction, including representation of action, affordance, pose, contact and physics [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Various examples showing how motion is added to the geometry. From left to right: [VNZ25, DSW∗ 25, LKLR24, QCZ∗ 23] 3. Modeling Motion The core of motion representation is establishing how geometry evolves across temporal frames. This can be formulated as deter￾mining the geometric state Gt as time t evolves. We distinguish four principal motion classes in 4D representations: articulated motion, deformatio… view at source ↗
Figure 5
Figure 5. Figure 5: Interaction specific representations. (a) Parametrizing poses between interaction entities [SDT∗ 22]; (b) Representing con￾tacts [HGT∗ 21]; (c) Representing affordances [DXW∗ 21]; (d) Representing physical properties [ZBS∗ 24]. the interaction signal—canonicalizing to a single reference frame discards this critical information. Independent of coordinate choice, the rotation parametrization itself requires … view at source ↗
Figure 6
Figure 6. Figure 6: Overview of key datasets in 4D representation research: (a) Objaverse-Dy [YXV∗ 25] provides a large-scale collection of synthetic animated 3D assets, serving as the primary source for current 4D generation works after filtering; (b) ActorsHQ [IRG∗ 23] is a high-fidelity, multi-view human performance capture dataset with per-frame 3D mesh reconstructions for dynamic human motion (c) Stereo4D [JTL∗ 24] repre… view at source ↗
Figure 7
Figure 7. Figure 7: Overview of three main training paradigms for 4D generation: (a) Per-scene optimization treats each scene as an individual opti￾mization problem, leveraging information distilled from large foundation models (e.g., image and video diffusion models via score distillation sampling) to optimize a 4D representation; (b) End-to-end training employs feed-forward models trained on large-scale native 4D assets to … view at source ↗

discussion (0)

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

Forward citations

Cited by 1 Pith paper

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

  1. One Video, One World: Turning Monocular Video into Physical 4D Scenes

    cs.CV 2026-06 unverdicted novelty 8.0

    OVOW reconstructs instance-level, simulation-ready 4D mesh scenes from monocular video via a four-stage training-free pipeline and introduces a new benchmark for structured Video-to-4D evaluation.

Reference graph

Works this paper leans on

300 extracted references · 1 canonical work pages · cited by 1 Pith paper

  1. [1]

    write newline

    " write newline "" before.all 'output.state := FUNCTION fin.entry.original add.period write newline FUNCTION new.block output.state before.all = 'skip after.block 'output.state := if FUNCTION new.sentence output.state after.block = 'skip output.state before.all = 'skip after.sentence 'output.state := if if FUNCTION not #0 #1 if FUNCTION and 'skip pop #0 i...

  2. [2]

    : Physically embodied gaussian splatting: Embedding physical priors into a visual 3d world model for robotics

    Abou-Chakra J., Rana K., Dayoub F., S \"u nderhauf N. : Physically embodied gaussian splatting: Embedding physical priors into a visual 3d world model for robotics. In Conference on Robot Learning (2023), no. 7th

    2023

  3. [3]

    https://www.mixamo.com, 2015

    Adobe : Mixamo. https://www.mixamo.com, 2015

    2015

  4. [4]

    : Hyperreel: High-fidelity 6-dof video with ray-conditioned sampling

    Attal B., Huang J.-B., Richardt C., Zollhoefer M., Kopf J., O’Toole M., Kim C. : Hyperreel: High-fidelity 6-dof video with ray-conditioned sampling. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (2023)

    2023

  5. [5]

    : Score distillation sampling with learned manifold corrective

    Alldieck T., Kolotouros N., Sminchisescu C. : Score distillation sampling with learned manifold corrective. In European Conference on Computer Vision (2024), Springer, pp. 1--18

    2024

  6. [6]

    : Re-imagine the negative prompt algorithm: Transform 2d diffusion into 3d, alleviate janus problem and beyond

    Armandpour M., Sadeghian A., Sadeghian A., Wang H., Hu C., Wang Z. : Re-imagine the negative prompt algorithm: Transform 2d diffusion into 3d, alleviate janus problem and beyond. arXiv preprint arXiv:2304.04968 (2023)

    Pith/arXiv arXiv 2023

  7. [7]

    D., Sminchisescu C

    Andriluka M., Tabanpour B., Freeman C. D., Sminchisescu C. : Learned neural physics simulation for articulated 3d human pose reconstruction. In European Conference on Computer Vision (2024), Springer, pp. 320--336

    2024

  8. [8]

    : Learning to generalize kinematic models to novel objects

    Abbatematteo B., Tellex S., Konidaris G. : Learning to generalize kinematic models to novel objects. In Proceedings of the 3rd Conference on Robot Learning (2019)

    2019

  9. [9]

    : A differentiable extended kalman filter for object tracking

    Bauza M., et al. : A differentiable extended kalman filter for object tracking. Frontiers in Robotics and AI (2021). URL: https://www.frontiersin.org/articles/10.3389/frobt.2021.686447

    arXiv 2021

  10. [10]

    Berlincioni L., Berretti S., Bertini M., Bimbo A. D. : 4dsr-gcn: 4d video point cloud upsampling using graph convolutional networks. In Proceedings of the 1st International Workshop on Multimedia Content Generation and Evaluation: New Methods and Practice (2023), pp. 57--65

    2023

  11. [11]

    : Stable video diffusion: Scaling latent video diffusion models to large datasets

    Blattmann A., Dockhorn T., Kulal S., Mendelevitch D., Kilian M., Lorenz D., Levi Y., English Z., Voleti V., Letts A., et al. : Stable video diffusion: Scaling latent video diffusion models to large datasets. arXiv preprint arXiv:2311.15127 (2023)

    Pith/arXiv arXiv 2023

  12. [12]

    C., Hays J

    Brahmbhatt S., Ham C., Kemp C. C., Hays J. : Contactdb: Analyzing and predicting grasp contact via thermal imaging. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition (2019), pp. 8709--8719

    2019

  13. [13]

    : Per-gaussian embedding-based deformation for deformable 3d gaussian splatting

    Bae J., Kim S., Yun Y., Lee H., Bang G., Uh Y. : Per-gaussian embedding-based deformation for deformable 3d gaussian splatting. arXiv preprint arXiv:2404.03613 (2024)

    Pith/arXiv arXiv 2024

  14. [14]

    : Frozen in time: A joint video and image encoder for end-to-end retrieval

    Bain M., Nagrani A., Varol G., Zisserman A. : Frozen in time: A joint video and image encoder for end-to-end retrieval. Proceedings of the IEEE/CVF International Conference on Computer Vision (2021), 1728--1738

    2021

  15. [15]

    : Interaction networks for learning about objects, relations and physics

    Battaglia P., Pascanu R., Lai M., Jimenez Rezende D., et al. : Interaction networks for learning about objects, relations and physics. Advances in neural information processing systems 29 (2016)

    2016

  16. [16]

    W., Fidler S., Kreis K

    Blattmann A., Rombach R., Ling H., Dockhorn T., Kim S. W., Fidler S., Kreis K. : Align your latents: High-resolution video synthesis with latent diffusion models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (2023)

    2023

  17. [17]

    J., Tagliasacchi A., Lindell D

    Bahmani S., Skorokhodov I., Rong V., Wetzstein G., Guibas L., Wonka P., Tulyakov S., Park J. J., Tagliasacchi A., Lindell D. B. : 4d-fy: Text-to-4d generation using hybrid score distillation sampling. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (2024)

    2024

  18. [18]

    B., Gojcic Z., Fidler S., et al

    Bahmani S., Shen T., Ren J., Huang J., Jiang Y., Turki H., Tagliasacchi A., Lindell D. B., Gojcic Z., Fidler S., et al. : Lyra: Generative 3d scene reconstruction via video diffusion model self-distillation. arXiv preprint arXiv:2509.19296 (2025)

    arXiv 2025

  19. [19]

    : Learning spatiotemporal graphs of human activities

    Brendel W., Todorovic S. : Learning spatiotemporal graphs of human activities. In 2011 International Conference on Computer Vision (2011), IEEE, pp. 778--785

    2011

  20. [20]

    : Lumiere: A space-time diffusion model for video generation

    Bar-Tal O., Chefer H., Tov O., Herrmann C., Paiss R., Zada S., Ephrat A., Hur J., Li Y., Michaeli T., et al. : Lumiere: A space-time diffusion model for video generation. arXiv preprint arXiv:2401.12945 (2024)

    Pith/arXiv arXiv 2024

  21. [21]

    D., Kemp C

    Brahmbhatt S., Tang C., Twigg C. D., Kemp C. C., Hays J. : Contactpose: A dataset of grasps with object contact and hand pose. In European Conference on Computer Vision (2020), Springer, pp. 361--378

    2020

  22. [22]

    L., Xie X., Petrov I., Sminchisescu C., Theobalt C., Pons-Moll G

    Bhatnagar B. L., Xie X., Petrov I., Sminchisescu C., Theobalt C., Pons-Moll G. : Behave: Dataset and method for tracking human object interactions. In IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (jun 2022), IEEE

    2022

  23. [23]

    : Object-centric reconstruction and tracking of dynamic unknown objects using 3d gaussian splatting

    Chen X., et al. : Object-centric reconstruction and tracking of dynamic unknown objects using 3d gaussian splatting. arXiv preprint arXiv:2405.20104 (2024)

    Pith/arXiv arXiv 2024

  24. [24]

    H., Vora S., Liong V

    Caesar H., Bankiti V., Lang A. H., Vora S., Liong V. E., Xu Q., Krishnan A., Pan Y., Baldan G., Beijbom O. : nuscenes: A multimodal dataset for autonomous driving. Proceedings of the IEEE/CVF conference on computer vision and pattern recognition (2020), 11621--11631

    2020

  25. [25]

    : Physx-3d: Physical-grounded 3d asset generation

    Cao Z., Chen Z., Pan L., Liu Z. : Physx-3d: Physical-grounded 3d asset generation. arXiv preprint arXiv:2507.12465 (2025)

    arXiv 2025

  26. [26]

    : Easi3r: Estimating disentangled motion from dust3r without training

    Chen X., Chen Y., Xiu Y., Geiger A., Chen A. : Easi3r: Estimating disentangled motion from dust3r without training. arXiv preprint arXiv:2503.24391 (2025)

    arXiv 2025

  27. [27]

    : Human3r: Everyone everywhere all at once

    Chen Y., Chen X., Xue Y., Chen A., Xiu Y., Gerard P.-M. : Human3r: Everyone everywhere all at once. arXiv preprint arXiv:2510.06219 (2025)

    arXiv 2025

  28. [28]

    : Neural surface reconstruction of dynamic scenes with monocular rgb-d camera

    Cai H., Feng W., Feng X., Wang Y., Zhang J. : Neural surface reconstruction of dynamic scenes with monocular rgb-d camera. NeurIPS 35 (2022)

    2022

  29. [29]

    X., Funkhouser T., Guibas L., Hanrahan P., Huang Q., Li Z., Savarese S., Savva M., Song S., Su H., et al

    Chang A. X., Funkhouser T., Guibas L., Hanrahan P., Huang Q., Li Z., Savarese S., Savva M., Song S., Su H., et al. : Shapenet: An information-rich 3d model repository. In Proceedings of the IEEE conference on computer vision and pattern recognition (2015), pp. 81--89

    2015

  30. [30]

    F., Dideriksen T., Arora H., Guillaumin M., Malik J

    Collins J., Goel S., Deng K., Luthra A., Xu L., Gundogdu E., Zhang X., Yago Vicente T. F., Dideriksen T., Arora H., Guillaumin M., Malik J. : Abo: Dataset and benchmarks for real-world 3d object understanding. CVPR (2022)

    2022

  31. [31]

    : Periodic vibration gaussian: Dynamic urban scene reconstruction and real-time rendering

    Chen Y., Gu C., Jiang J., Zhu X., Zhang L. : Periodic vibration gaussian: Dynamic urban scene reconstruction and real-time rendering. ArXiv:2311.18561 (2023)

    Pith/arXiv arXiv 2023

  32. [32]

    : Ct4d: Consistent text-to-4d generation with animatable meshes

    Chen C., Huang S., Chen X., Chen G., Han X., Zhang K., Gong M. : Ct4d: Consistent text-to-4d generation with animatable meshes. arXiv preprint arXiv:2408.08342 (2024)

    Pith/arXiv arXiv 2024

  33. [33]

    : Occlusion-robust scene flow-based tissue deformation recovery incorporating a mesh optimization model

    Chen J., Hara K., Kobayashi E., Sakuma I., Tomii N. : Occlusion-robust scene flow-based tissue deformation recovery incorporating a mesh optimization model. International Journal of Computer Assisted Radiology and Surgery 18, 6 (2023), 1043--1051

    2023

  34. [34]

    R., Xu G., Ma F., Li M., Zhou G

    Chen G., He Y., Yu M., Yu F. R., Xu G., Ma F., Li M., Zhou G. : Inter3d: A benchmark and strong baseline for human-interactive 3d object reconstruction. arXiv preprint arXiv:2502.14004 (2025)

    Pith/arXiv arXiv 2025

  35. [35]

    : Hexplane: A fast representation for dynamic scenes

    Cao A., Johnson J. : Hexplane: A fast representation for dynamic scenes. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (2023)

    2023

  36. [36]

    : D-grasp: Physically plausible dynamic grasp synthesis for hand-object interactions

    Christen S., Kocabas M., Aksan E., Hwangbo J., Song J., Hilliges O. : D-grasp: Physically plausible dynamic grasp synthesis for hand-object interactions. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (2022), pp. 20577--20586

    2022

  37. [37]

    : Reconstructing 4d spatial intelligence: A survey, 2025

    Cao Y., Lu J., Huang Z., Shen Z., Zhao C., Hong F., Chen Z., Li X., Wang W., Liu Y., Liu Z. : Reconstructing 4d spatial intelligence: A survey, 2025. URL: https://arxiv.org/abs/2507.21045, http://arxiv.org/abs/2507.21045 arXiv:2507.21045

    Pith/arXiv arXiv 2025

  38. [38]

    : Motion2vecsets: 4d latent vector set diffusion for non-rigid shape reconstruction and tracking

    Cao W., Luo C., Zhang B., Nie ner M., Tang J. : Motion2vecsets: 4d latent vector set diffusion for non-rigid shape reconstruction and tracking. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition (2024), pp. 20496--20506

    2024

  39. [39]

    : 4dnex: Feed-forward 4d generative modeling made easy

    Chen Z., Liu T., Zhuo L., Ren J., Tao Z., Zhu H., Hong F., Pan L., Liu Z. : 4dnex: Feed-forward 4d generative modeling made easy. arXiv preprint arXiv:2508.13154 (2025)

    Pith/arXiv arXiv 2025

  40. [40]

    Y., Lee K

    Choi H., Moon G., Chang J. Y., Lee K. M. : Beyond static features for temporally consistent 3d human pose and shape from a video. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition (2021), pp. 1964--1973

    2021

  41. [41]

    : Ganhand: Predicting human grasp affordances in multi-object scenes

    Corona E., Pumarola A., Alenya G., Moreno-Noguer F., Rogez G. : Ganhand: Predicting human grasp affordances in multi-object scenes. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition (2020), pp. 5031--5041

    2020

  42. [42]

    K., Liu Z

    Cao Y., Pan L., Han K., Wong K.-Y. K., Liu Z. : Avatargo: Zero-shot 4d human-object interaction generation and animation. In The Thirteenth International Conference on Learning Representations (2024)

    2024

  43. [43]

    M., Kim S

    Chignoli M., Slotine J.-J., Wensing P. M., Kim S. : Urdf+: An enhanced urdf for robots with kinematic loops. In 2024 IEEE-RAS 23rd International Conference on Humanoid Robots (Humanoids) (2024), IEEE, pp. 197--204

    2024

  44. [44]

    : Urdformer: A pipeline for constructing articulated simulation environments from real-world images

    Chen Z., Walsman A., Memmel M., Mo K., Fang A., Vemuri K., Wu A., Fox D., Gupta A. : Urdformer: A pipeline for constructing articulated simulation environments from real-world images. arXiv preprint arXiv:2405.11656 (2024)

    Pith/arXiv arXiv 2024

  45. [45]

    M., Ivanovic B., Litany O., Gojcic Z., Fidler S., Pavone M., Song L., Wang Y

    Chen Z., Yang J., Huang J., de Lutio R., Esturo J. M., Ivanovic B., Litany O., Gojcic Z., Fidler S., Pavone M., Song L., Wang Y. : OmniRe : Omni Urban Scene Reconstruction , 2024. https://doi.org/10.48550/arXiv.2408.16760 doi:10.48550/arXiv.2408.16760

    Pith/arXiv arXiv doi:10.48550/arxiv.2408.16760 2024

  46. [46]

    : Neural-abc: neural parametric models for articulated body with clothes

    Chen H., Yao Y., Zhang J. : Neural-abc: neural parametric models for articulated body with clothes. IEEE Transactions on Visualization and Computer Graphics 31, 2 (2024), 1478--1495

    2024

  47. [47]

    : Deepverse: 4d autoregressive video generation as a world model

    Chen J., Zhu H., He X., Wang Y., Zhou J., Chang W., Zhou Y., Li Z., Fu Z., Pang J., et al. : Deepverse: 4d autoregressive video generation as a world model. arXiv preprint arXiv:2506.01103 (2025)

    Pith/arXiv arXiv 2025

  48. [48]

    : Physics-driven local-whole elastic deformation modeling for point cloud representation learning

    Chen Z., Zhao R., Han X., Guo X., Wang S., Qiao Z. : Physics-driven local-whole elastic deformation modeling for point cloud representation learning. arXiv preprint arXiv:2505.13812 (2025)

    Pith/arXiv arXiv 2025

  49. [49]

    : Encoder-decoder with atrous separable convolution for semantic image segmentation

    Chen L.-C., Zhu Y., Papandreou G., Schroff F., Adam H. : Encoder-decoder with atrous separable convolution for semantic image segmentation. In ECCV (2018)

    2018

  50. [50]

    : V2m4: 4d mesh animation reconstruction from a single monocular video

    Chen J., Zhang B., Tang X., Wonka P. : V2m4: 4d mesh animation reconstruction from a single monocular video. arXiv preprint arXiv:2503.09631 (2025)

    Pith/arXiv arXiv 2025

  51. [51]

    : Imagenet: A large-scale hierarchical image database

    Deng J., Dong W., Socher R., Li L.-J., Li K., Fei-Fei L. : Imagenet: A large-scale hierarchical image database. In 2009 IEEE conference on computer vision and pattern recognition (2009), Ieee, pp. 248--255

    2009

  52. [52]

    B., Vanhoucke V

    Downs L., Francis A., Koenig N., Kinman B., Hickman R., Reymann K., McHugh T. B., Vanhoucke V. : Google scanned objects: A high-quality dataset of 3d scanned household items. 2022 International Conference on Robotics and Automation (ICRA) (2022), 2553--2560

    2022

  53. [53]

    Y., et al

    Deitke M., Liu R., Wallingford M., Ngo H., Michel O., Kusupati A., Fan A., Laforte C., Voleti V., Gadre S. Y., et al. : Objaverse-xl: A universe of 10m+ 3d objects. In Advances in Neural Information Processing Systems (2023), vol. 36, pp. 35799--35813

    2023

  54. [54]

    P., Mandi Z., Yao Y., Liu J.-W., Seidenschwarz J., Shou M

    Duisterhof B. P., Mandi Z., Yao Y., Liu J.-W., Seidenschwarz J., Shou M. Z., Ramanan D., Song S., Birchfield S., Wen B., et al. : Deformgs: Scene flow in highly deformable scenes for deformable object manipulation. ArXiv:2312.00583 (2023)

    Pith/arXiv arXiv 2023

  55. [55]

    P., Mandi Z., Yao Y., Liu J.-W., Shou M

    Duisterhof B. P., Mandi Z., Yao Y., Liu J.-W., Shou M. Z., Song S., Ichnowski J. : Md-splatting: Learning metric deformation from 4d gaussians in highly deformable scenes

  56. [56]

    : Textmesh4d: High-quality text-to-4d mesh generation

    Dai S., Su X., Wan B., Hu R., Xu K. : Textmesh4d: High-quality text-to-4d mesh generation. arXiv preprint arXiv:2506.24121 (2025)

    Pith/arXiv arXiv 2025

  57. [57]

    : Scenefun3d: Fine-grained functionality and affordance understanding in 3d scenes

    Delitzas A., Takmaz A., Tombari F., Sumner R., Pollefeys M., Engelmann F. : Scenefun3d: Fine-grained functionality and affordance understanding in 3d scenes. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (2024), pp. 14531--14542

    2024

  58. [58]

    : 4d-rotor gaussian splatting: towards efficient novel view synthesis for dynamic scenes

    Duan Y., Wei F., Dai Q., He Y., Chen W., Chen B. : 4d-rotor gaussian splatting: towards efficient novel view synthesis for dynamic scenes. In ACM SIGGRAPH 2024 Conference Papers (2024), pp. 1--11

    2024

  59. [59]

    : 3d affordancenet: A benchmark for visual object affordance understanding

    Deng S., Xu X., Wu C., Chen K., Jia K. : 3d affordancenet: A benchmark for visual object affordance understanding. In proceedings of the IEEE/CVF conference on computer vision and pattern recognition (2021), pp. 1778--1787

    2021

  60. [60]

    : Nsm4d: Neural scene model based online 4d point cloud sequence understanding

    Dong Y., Zhang Z., Liu Y., Yi L. : Nsm4d: Neural scene model based online 4d point cloud sequence understanding. arXiv preprint arXiv:2310.08326 (2023)

    Pith/arXiv arXiv 2023

  61. [61]

    B., Wu J

    Du Y., Zhang Y., Yu H.-X., Tenenbaum J. B., Wu J. : Neural radiance flow for 4d view synthesis and video processing. In 2021 IEEE/CVF International Conference on Computer Vision (ICCV) (2021), IEEE Computer Society, pp. 14304--14314

    2021

  62. [62]

    R., Zhou Y., et al

    Ettinger S., Cheng S., Caine B., Liu C., Zhao H., Pradhan S., Chai Y., Sapp B., Qi C. R., Zhou Y., et al. : Large scale interactive motion forecasting for autonomous driving: The waymo open motion dataset. In Proceedings of the IEEE/CVF international conference on computer vision (2021), pp. 9710--9719

    2021

  63. [63]

    : Flowbot3d: Learning 3d articulation flow to manipulate articulated objects

    Eisner B., Zhang H., Held D. : Flowbot3d: Learning 3d articulation flow to manipulate articulated objects. arXiv preprint arXiv:2205.04382 (2022)

    Pith/arXiv arXiv 2022

  64. [64]

    : 3d human interaction generation: A survey

    Fan S., Huang W., Cai X., Du B. : 3d human interaction generation: A survey. arXiv preprint arXiv:2503.13120 (2025)

    Pith/arXiv arXiv 2025

  65. [65]

    : 3d-future: 3d furniture shape with texture

    Fu H., Jia R., Gao L., Gong M., Zhao B., Maybank S., Tao D. : 3d-future: 3d furniture shape with texture. International Journal of Computer Vision 129, 12 (2021), 3313--3337

    2021

  66. [66]

    R., Porzi L., Pollefeys M., Kontschieder P

    Fischer T., Kulhanek J., Bul \`o S. R., Porzi L., Pollefeys M., Kontschieder P. : Dynamic 3d gaussian fields for urban areas. In NeurIPS (2024)

    2024

  67. [67]

    R., Recht B., Kanazawa A

    Fridovich-Keil S., Meanti G., Warburg F. R., Recht B., Kanazawa A. : K-planes: Explicit radiance fields in space, time, and appearance. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (2023)

    2023

  68. [68]

    R., Pollefeys M., Kontschieder P

    Fischer T., Porzi L., Bulo S. R., Pollefeys M., Kontschieder P. : Multi-level neural scene graphs for dynamic urban environments. In CVPR (2024)

    2024

  69. [69]

    : Pie-nerf: Physics-based interactive elastodynamics with nerf

    Feng Y., Shang Y., Li X., Shao T., Jiang C., Yang Y. : Pie-nerf: Physics-based interactive elastodynamics with nerf. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (2024), pp. 4450--4461

    2024

  70. [70]

    : Fast dynamic radiance fields with time-aware neural voxels

    Fang J., Yi T., Wang X., Xie L., Zhang X., Liu W., Nie ner M., Tian Q. : Fast dynamic radiance fields with time-aware neural voxels. In SIGGRAPH Asia 2022 Conference Papers (2022), pp. 1--9

    2022

  71. [71]

    : Three-dimensional reconstruction of human interactions

    Fieraru M., Zanfir M., Oneata E., Popa A.-I., Olaru V., Sminchisescu C. : Three-dimensional reconstruction of human interactions. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (2020), pp. 7214--7223

    2020

  72. [72]

    J., Darrell T., Kanazawa A

    Feng H., Zhang J., Wang Q., Ye Y., Yu P., Black M. J., Darrell T., Kanazawa A. : St4rtrack: Simultaneous 4d reconstruction and tracking in the world. arXiv preprint arXiv:2504.13152 (2025)

    Pith/arXiv arXiv 2025

  73. [73]

    : Advances in radiance field for dynamic scene: From neural field to gaussian field, 2025

    Fan J., Zeng X., Zhang J., Gong M., Yang Y., Tao D. : Advances in radiance field for dynamic scene: From neural field to gaussian field, 2025. URL: https://arxiv.org/abs/2505.10049, http://arxiv.org/abs/2505.10049 arXiv:2505.10049

    Pith/arXiv arXiv 2025

  74. [74]

    : Neural deformable voxel grid for fast optimization of dynamic view synthesis

    Guo X., Chen G., Dai Y., Ye X., Sun J., Tan X., Ding E. : Neural deformable voxel grid for fast optimization of dynamic view synthesis. In ACCV (2022)

    2022

  75. [75]

    K., Gopalan N

    Gupta A., Gu W., Patil O., Lee J. K., Gopalan N. : Learning sequential kinematic models from demonstrations for multi-jointed articulated objects. arXiv preprint arXiv:2505.06363 (2025)

    Pith/arXiv arXiv 2025

  76. [76]

    T., Poole B

    Gao R., Holynski A., Henzler P., Brussee A., Martin-Brualla R., Srinivasan P., Barron J. T., Poole B. : Cat3d: Create anything in 3d with multi-view diffusion models. arXiv preprint arXiv:2405.10314 (2024)

    Pith/arXiv arXiv 2024

  77. [77]

    Gibson J. J. : The theory of affordances. Perceiving, acting, and knowing: toward an ecological psychology (1977), pp--67

    1977

  78. [78]

    : Vid2avatar: 3d avatar reconstruction from videos in the wild via self-supervised scene decomposition

    Guo C., Jiang T., Chen X., Song J., Hilliges O. : Vid2avatar: 3d avatar reconstruction from videos in the wild via self-supervised scene decomposition. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (2023)

    2023

  79. [79]

    : Vision meets robotics: The kitti dataset

    Geiger A., Lenz P., Stiller C., Urtasun R. : Vision meets robotics: The kitti dataset. The International Journal of Robotics Research 32, 11 (2013), 1231--1237

    2013

  80. [80]

    D., Kalogerakis E

    Goyal P., Petrov D., Andrews S., Ben-Shabat Y., Liu H.-T. D., Kalogerakis E. : Geopard: Geometric pretraining for articulation prediction in 3d shapes. arXiv preprint arXiv:2504.02747 (2025)

    Pith/arXiv arXiv 2025

Showing first 80 references.