Pith. sign in

REVIEW 4 major objections 5 minor 3 cited by

4D-LRM: Large Space-Time Reconstruction Model From and To Any View at Any Time

T0 review · 4 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read 4D-LRM claims that a single feed-forward transformer can reconstruct a dynamic object from a few posed images at scattered times and render any view at any time, in under 1.5 seconds per 24-frame sequence.

desk verdict A real step forward for feed-forward 4D reconstruction, with an honest limitations section; the "any view at any time" headline overstates what the affine-in-time Gaussian representation and synthetic-only evaluation can support. read the letter →

arxiv 2506.18890 v1 pith:L35SFJBE submitted 2025-06-23 cs.CV

Ziqiao Ma , Xuweiyi Chen , Shoubin Yu , Sai Bi , Kai Zhang , Chen Ziwen , Sihan Xu , Jianing Yang
show 5 more authors
Zexiang Xu Kalyan Sunkavalli Mohit Bansal Joyce Chai Hao Tan
This is my paper · ORCID
classification cs.CV
keywords 4DreconstructionlargemodelGaussiansplattingnovelview-timesynthesisspatiotemporalpretrainingdynamicobjectfeed-forwardtransformerframeinterpolation
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper argues that 4D reconstruction—rebuilding a moving object from a few images—can be treated as a large-scale pretraining problem rather than per-scene optimization. Its model, 4D-LRM, takes any number of posed images at arbitrary timestamps and, in one transformer forward pass, predicts a cloud of 4D Gaussians that can render any novel view at any novel time. The authors claim this is the first large-scale 4D reconstruction model with unconstrained spatiotemporal inputs, and report that a 24-frame sequence is reconstructed in under 1.5 seconds on a single A100 GPU, with PSNR above 30 on structured camera setups. If true, this would make dynamic-object reconstruction fast enough for practical 4D asset generation, AR/VR, and robotics pipelines, replacing slow optimization or unreliable generative synthesis.

What carries the argument

The load-bearing object is the 4D Gaussian primitive: a 20-dimensional parameter vector with RGB color, four scales, two quaternions for 4D rotation, opacity, and a space-time center (x, y, z, t). Its covariance is a 4D ellipsoid, and Eq. (11) gives the conditional 3D Gaussian at any time $t$ by a closed-form marginalization and conditioning of the multivariate normal; the tile-based Gaussian rasterizer then splats these conditional Gaussians. A transformer with Plücker-ray and timestamp tokenization decodes one such primitive per pixel, optionally joined by learnable free Gaussian tokens for sparse-view or generative setups. This machinery is what lets the model share information across both views and time in a single forward pass and render at effectively continuous frame rates.

What would settle it

Take a held-out dynamic object whose motion is deliberately non-linear or includes occlusion (e.g., a swinging cloth or a closing hand), render sparse posed views around it, and run the released 4D-LRM: if PSNR against ground truth drops substantially and ghosting or motion trails appear, the central claim is false for such motions; if the render stays sharp, the representation generalizes beyond smooth trajectories.

Watch

Extended reading notes

Core claim

The central claim is that a unified space-time representation, in which every primitive is a 4D Gaussian with a spatial and temporal center, lets a feed-forward transformer learn a general prior over dynamic objects. Given posed image tokens augmented with ray geometry and a timestamp, the model directly regresses per-pixel 4D Gaussian parameters; at render time each 4D Gaussian is marginalized in time and conditioned to a 3D Gaussian at the queried instant via Eq. (11). The paper reports that this representation, trained on rendered synthetic objects, generalizes to novel objects, interpolates missing timestamps, and handles diverse camera setups, matching or beating per-frame 3D reconstruction baselines when views are sparse, and serving as a fast, faithful backbone for 4D generation.

Load-bearing premise

The load-bearing premise is that a moving object can be divided into smooth, ellipsoidal, locally linear pieces, so that each 4D Gaussian with a single time-variance captures one chunk of motion; the model's own failure case shows this premise breaks for non-linear trajectories and occlusion, producing ghosting artifacts.

Editorial extensions

If this is right

  • A 24-frame dynamic object can be reconstructed in one forward pass in under 1.5 seconds on one A100 GPU, and rendering can be done at arbitrary intermediate timestamps, in principle at infinite frame rate.
  • The model generalizes to novel objects and to unconstrained camera setups, including random views per frame and missing half the timestamps, with PSNR above 30 in the easier canonical setups.
  • Scaling model size and training data improves reconstruction quality, and initializing from a static-scene LRM accelerates convergence.
  • With free Gaussian tokens and a generative prior for the first frame, the same model exceeds generation-based 4D methods in faithfulness and speed on held-out 4D benchmarks.
  • Performance improves with more input views up to about 48 views, then declines slightly because of overcrowded Gaussians and long input sequences.

Reading between the lines

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

  • A reader might infer that extending the same architecture with long-context handling and test-time training—both named by the paper as future work—could push toward hundreds of high-resolution inputs without changing the core representation.
  • A testable extension is to train or fine-tune 4D-LRM on real in-the-wild dynamic video with estimated poses; the paper's smooth-ellipsoid motion assumption predicts visible ghosting on fast, non-rigid, or occluded motion, so real-world evaluation would show whether the synthetic-data prior transfers.
  • The 4D Gaussian representation ties reconstruction quality to trajectory smoothness; for articulated or branching motion, one could hybridize with deformation fields or trajectory splines, though the paper does not explore this.
  • Since static 3D reconstruction with time collapsed to zero still gives usable quality, the same model could serve as a single backbone for both 3D and 4D tasks.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The paper introduces 4D-LRM, a transformer-based feed-forward model that takes sparse posed images with timestamps and predicts per-pixel 4D Gaussian primitives, adapting the 4DGS representation of Yang et al. The model is trained on a curated subset of Objaverse (animated objects from Diffusion4D, plus static objects with small linear displacements) and evaluated by re-rendering Consistent4D and a held-out Objaverse4D split. The paper reports PSNR above 30 on several structured camera setups, strong performance relative to GS-LRM and SoM, 24-frame reconstruction in less than 1.5 seconds on a single A100, scaling analyses, and an application to 4D generation by chaining with SV3D. The authors claim this is the first large-scale 4D reconstruction model that accepts unconstrained view-time inputs and renders arbitrary novel view-time combinations.

Significance. If the empirical claims hold, this is a meaningful step for feed-forward 4D reconstruction: a single model that jointly reasons over space and time, generalizes to held-out objects, interpolates across time, and handles diverse camera configurations. The strengths of the paper are its clear formulation of the unified space-time representation, the breadth of camera setups tested, the held-out evaluation protocol, the inclusion of external baselines, and a candid failure analysis. At the same time, the central claim is currently scoped too broadly: the representation itself can only express affine-in-time, constant-shape primitives, and all dynamic evaluations are synthetic re-renders, so the regime where the headline claim holds is not yet measured. The manuscript also lacks error bars on all main quantitative tables, which is important because several reported differences are small or non-monotonic.

major comments (4)
  1. [§2.1, Eq. (11); App. B.2, Fig. 11] The central claim of arbitrary novel view-time combinations is bounded by the 4DGS representation used. In Eq. (11), the conditional mean μ_{xyz|t} is affine in t and the conditional covariance is independent of t, so each primitive can only follow a straight-line, constant-shape trajectory. The paper's own failure analysis in App. B.2 and Fig. 11 confirms that non-linear trajectories and self-occlusion produce temporal ghosting and motion trails. Since all Table 1–2 evaluations are synthetic re-renders and no setup is stratified by motion complexity or occlusion, the regime in which 'any view at any time' holds is not measured. Please either add a test split that varies motion curvature and occlusion (or a real dynamic video benchmark) and report performance there, or restrict the headline claim to smooth, well-covered motions.
  2. [Tables 1–3] All quantitative claims rest on single-run point estimates without error bars, repeated seeds, or significance tests. Several reported comparisons are close or non-monotonic, e.g., Table 3 shows 4D-LRM at 28.343 dB for 1 random view per frame versus 27.940 dB for 4 random views per frame, and Table 1 shows 128-resolution results that are sometimes higher than 256-resolution results for the same model. Without variance estimates it is hard to know whether the reported ordering is robust. Please report means with standard deviations over at least three seeds or a bootstrap over test objects for the main tables.
  3. [§3.2–3.3; §B.1] The evaluation contains no real-world dynamic scene. All 4D benchmarks are re-renders of synthetic Objaverse and Consistent4D assets, and the only real-image evaluation is the static GSO dataset in Table 5. Because the contribution is a learned general 4D prior, the paper needs at least a qualitative or quantitative evaluation on real dynamic videos (with the resulting limitations on segmentation, masks, and non-synthetic motion) to support the claim of general 4D reconstruction. Without this, the central result is only demonstrated for synthetic object videos.
  4. [§3.1, App. A.3] The training data relies on the assumption that static Objaverse objects with minor frame-by-frame displacements along a single random direction are useful dynamic training data, and this assumption is not isolated in any ablation. The mixing ratio HQ4D:4D:static is fixed at 200:50:1, so the model's ability to handle arbitrary dynamics is only shown for the limited motion distribution of the training set. Please add an ablation or a test set with dynamics substantially outside the training distribution (e.g., articulated or occluded motions, non-linear paths) to support the scaling and generalization claims.
minor comments (5)
  1. [§3.1] The text says the evaluation re-renders the first 48 (2×24) frames of Consistent4D, but Tables 1–2 and Figure 4 discuss 24-frame sequences; please clarify the exact frame range used in each table and figure.
  2. [Figure 3] The architecture diagram includes Free Gaussian Tokens with dimension N, but the default pretraining uses no free Gaussians; please mark this component as optional and state N=0 for the main reconstruction experiments.
  3. [Eq. (3) and App. A.1] The perceptual loss weight λ is set to 0.5 and the appendix says the relevant hyperparameters are 'relatively insensitive,' but no sensitivity study is shown; a short plot or table would make this claim verifiable.
  4. [Table 3] For the 'Canon. 4' row, please state the total number of input images (4 views per frame × 24 frames = 96) and clarify whether the random-view rows use the same total input budget or the same number of views per frame.
  5. [App. A.1] The notation for the decoded Gaussian parameters is split across Section 2.2, Algorithm 1, and App. A.1; unifying the names (e.g., g_scale vs. scale_xyz) and stating the depth clipping range [−1,1]^3 in the main text would improve readability.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: 4D-LRM's predictions are held-out renderings, and self-citations are limited to initialization and baseline choices.

full rationale

The paper's load-bearing derivation is a feed-forward image-to-4DGS predictor: posed images are tokenized with Plücker rays and timestamps, processed by a Transformer, and decoded into per-pixel 4D Gaussian primitives, which are rendered with the 4DGS rasterizer and trained with MSE plus perceptual loss against held-out supervision views. No predicted quantity is obtained by fitting a parameter to the test targets; evaluations are on held-out objects from Consistent4D and the Objaverse4D test split, against external baselines such as GS-LRM, SoM, L4GM, and SV4D. The 4DGS conditional-Gaussian equations (Eq. 2 and Eq. 11) are a mathematical consequence of conditioning a multivariate Gaussian, not an encoding of the test answer. The use of the authors' own GS-LRM for initialization and as a comparison baseline is a normal engineering practice, and the paper also reports from-scratch results, so the central spatiotemporal generalization claim does not reduce to that self-citation. The acknowledged limitation in Appendix B.2, that the ellipsoidal and locally linear Gaussian representation struggles with non-linear trajectories and occlusion, is a scoped correctness limitation of the adopted representation, not a circular step: the model is still tested on held-out data and its failures are empirically documented. Thus the paper is self-contained against external benchmarks and exhibits no significant circularity.

Assumptions & free parameters 6 free parameters · 5 assumptions · 1 invented entities

The method does not derive a physical law; its claims rest on learned model weights, hand-set hyperparameters, and a synthetic training distribution. The most consequential assumptions are that 4D Gaussians can express the target motions and that synthetic object renders transfer beyond the training distribution. No new physical entities are proposed, apart from the internal free Gaussian token mechanism.

free parameters (6)
  • Depth bounds delta_near and delta_far = 0.1 and 4.5
    Map sigmoid output to ray distance in the pixel-aligned Gaussian decoder, carried from GS-LRM [92] and not learned from 4D-LRM data.
  • Scale and opacity activation biases and clipping = scale bias -2.3, spatial max 0.3, temporal max 1.0, opacity bias -2.0
    Hand-chosen constants in Appendix A.1 to stabilize training; the paper states performance is relatively insensitive to their values.
  • Perceptual loss weight lambda = 0.5
    Set empirically in Eq. (3) for the reconstruction objective.
  • Dataset mixing ratio HQ4D:4D:static = 200:50:1
    Chosen for pretraining in Appendix A.3; it determines the distribution of motion and shape priors the model learns.
  • Image tokenizer patch size = 8x8
    Architecture choice in Appendix A.4; affects token count, context length, and the number of pixel-aligned Gaussians.
  • Curriculum training schedule = 128x128 for 100k steps, then 256x256 for 20k steps
    Computational cost tradeoff in Section 3.1; not fitted to a benchmark but influences final performance.
assumptions (5)
  • standard math The conditional distribution of a multivariate Gaussian is Gaussian, so the 4DGS sampling equations in Eq. (2) and Eq. (11) are valid.
    Standard Gaussian identities are used without proof, cited to 4DGS [82].
  • domain assumption Input camera poses and timestamps are known exactly and correctly synchronized across views.
    The method encodes Plücker rays and a timestamp map as input features, so pose or timestamp errors would corrupt the learned space-time correspondence.
  • domain assumption Rendered synthetic object videos from Objaverse4D are a sufficient training distribution for the claimed generalization to novel objects and downstream applications.
    All headline evaluations use re-rendered synthetic objects; no real-world dynamic video benchmark is included, as noted in Sections 3.2 and 6.
  • domain assumption A cloud of 4D Gaussians with smooth ellipsoidal support can represent arbitrary dynamic motion.
    The model inherits this from 4DGS [82], and the paper's Figure 11 shows non-linear trajectories are not captured efficiently.
  • ad hoc to paper Static Objaverse objects with small linear frame-by-frame displacements are useful dynamic training data.
    This augmentation, described in Section 3.1 and Appendix A.3, adds 783,000 static objects as pseudo-dynamic sequences; no theoretical or empirical justification is provided beyond training scale.
invented entities (1)
  • Free Gaussian tokens
    purpose: Learned tokens appended to image tokens to produce freeform 4D Gaussians not tied to pixels, used when view coverage is sparse in 4D generation fine-tuning.
    No external falsifiable handle exists for these tokens; they are trainable parameters whose benefit is measured only in ablations and the generation task.

how reviews work

0 comments
Cite this review

Pith. "Pith review of 4D-LRM: Large Space-Time Reconstruction Model From and To Any View at Any Time." pith.science (2026). https://pith.science/paper/L35SFJBE

@misc{pith2026250618890,
  author       = {Pith},
  title        = {Pith review of: 4D-LRM: Large Space-Time Reconstruction Model From and To Any View at Any Time},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/L35SFJBE}},
  note         = {Machine review of arXiv:2506.18890}
}
read the original abstract

Can we scale 4D pretraining to learn general space-time representations that reconstruct an object from a few views at some times to any view at any time? We provide an affirmative answer with 4D-LRM, the first large-scale 4D reconstruction model that takes input from unconstrained views and timestamps and renders arbitrary novel view-time combinations. Unlike prior 4D approaches, e.g., optimization-based, geometry-based, or generative, that struggle with efficiency, generalization, or faithfulness, 4D-LRM learns a unified space-time representation and directly predicts per-pixel 4D Gaussian primitives from posed image tokens across time, enabling fast, high-quality rendering at, in principle, infinite frame rate. Our results demonstrate that scaling spatiotemporal pretraining enables accurate and efficient 4D reconstruction. We show that 4D-LRM generalizes to novel objects, interpolates across time, and handles diverse camera setups. It reconstructs 24-frame sequences in one forward pass with less than 1.5 seconds on a single A100 GPU.

Figures

Figures reproduced from arXiv: 2506.18890 by the authors.

Figure 1
Figure 1. Large Space-Time Reconstruction Model (4D-LRM) is a data-driven 4D reconstruction model that takes sparse input views at any time and renders arbitrary novel view-time combinations. Abstract Can we scale 4D pretraining to learn general space-time representations that recon￾struct an object from a few views at some times to any view at any time? We provide an affirmative answer with 4D-LRM, the first large-scale 4D r… view at source ↗
Figure 2
Figure 2. Comparison between previous generative 4D modeling methods (e.g., L4GM [ [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Overview of 4D-LRM. 4D-LRM adopts a unified treatment of space and time, representing a dynamic object as a cloud of anisotropic 4D Gaussians [82]. We train a simple Transformer to regress 4D Gaussian primitives from a set of images with camera poses and timestamps. Each input image is tokenized by patchifying the temporally posed frames. The resulting multi-view image tokens are concatenated in temporal order and p… view at source ↗
Figures from the paper (10 more)
Figure 4
Figure 4. Figure 4: Different types of camera setups in evaluations: [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: Visual comparison with GS-LRM [92] using (a) all input views across time and (b) three random views from the same timestamp. 4D-LRM reconstructs novel view-time combinations by learning spatiotemporal representations from sparse inputs, outperforming per-frame reconstr…
Figure 6
Figure 6. Figure 6: Qualitative examples of 4D-LRM taking views with missing frames, including [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 7
Figure 7. Figure 7: Visual comparisons of 4D-LRM(-Free) to generation-based 4D models. For fair comparisons, [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]
Figure 8
Figure 8. Figure 8: We visualize the distributions of µt and Σt under Alternating Canonical Views and Frame Interpolation setups on 24 frames with the same dynamic object. 10000 20000 40000 60000 80000 2500 5000 7500 Training Steps 18 20 22 24 26 PSNR ( ) 4D-LRM-base #Target x 2 w/ HexPla…
Figure 9
Figure 9. Figure 9: Training-time scaling curves. Tested on Consistent4D (re-rendered). We compute the [PITH_FULL_IMAGE:figures/full_fig_p009_9.png]
Figure 10
Figure 10. Figure 10: Inference-time scaling curves. Tested on Consistent4D (re-rendered). We compute the [PITH_FULL_IMAGE:figures/full_fig_p010_10.png]
Figure 11
Figure 11. Figure 11: A typical failure case as 4D-LRM sometimes struggles with non-linear motion trajectories. [PITH_FULL_IMAGE:figures/full_fig_p022_11.png]
Figure 12
Figure 12. Figure 12: Qualitative examples of 4D-LRM under varying camera setups. We show the performance [PITH_FULL_IMAGE:figures/full_fig_p023_12.png]
Figure 13
Figure 13. Figure 13: Additional frame interpolation examples. We insert 4 [PITH_FULL_IMAGE:figures/full_fig_p024_13.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 3 Pith papers

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

  1. 4DHumanDiff: Direct Text-to-4DGS Generation for Consistent 360-Degree Dynamic Humans

    cs.CV 2026-07 conditional novelty 6.0 of 10

    A diffusion model trained on 60,000 fitted 4D Gaussian Splatting human clips generates text-prompted, view-consistent dynamic humans directly in 4D, over 10x faster than video-first pipelines.

  2. LSRM: High-Fidelity Object-Centric Reconstruction via Scaled Context Windows

    cs.CV 2026-04 conditional novelty 6.0 of 10

    Scaling transformer context with sparse attention and 3D-aware block routing improves feed-forward 3D reconstruction and inverse rendering, closing much of the quality gap with dense-view optimization.

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

    cs.CV 2025-10 conditional novelty 4.0 of 10

    A representation-centric survey of 4D generation and reconstruction, organized by geometry, motion, and interaction, with qualitative trade-off comparisons across seven representation families.

Reference graph

Works this paper leans on

98 extracted references · 49 canonical work pages · cited by 3 Pith papers

  1. [1]

    Building rome in a day.Communications of the ACM, 54(10):105–112, 2011

    Sameer Agarwal, Yasutaka Furukawa, Noah Snavely, Ian Simon, Brian Curless, Steven M Seitz, and Richard Szeliski. Building rome in a day.Communications of the ACM, 54(10):105–112, 2011

    2011

  2. [2]

    Layer normalization.arXiv preprint arXiv:1607.06450, 2016

    Jimmy Lei Ba, Jamie Ryan Kiros, and Geoffrey E Hinton. Layer normalization.arXiv preprint arXiv:1607.06450, 2016

    arXiv 2016

  3. [3]

    Hexplane: Afastrepresentationfordynamicscenes

    AngCaoandJustinJohnson. Hexplane: Afastrepresentationfordynamicscenes. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 130–141, 2023

    2023

  4. [4]

    Efficient geometry- aware 3d generative adversarial networks

    Eric R Chan, Connor Z Lin, Matthew A Chan, Koki Nagano, Boxiao Pan, Shalini De Mello, Orazio Gallo, Leonidas J Guibas, Jonathan Tremblay, Sameh Khamis, et al. Efficient geometry- aware 3d generative adversarial networks. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 16123–16133, 2022

    2022

  5. [5]

    Hardware-constrained hybrid coding of video imagery

    Luen C Chan and Peter Whiteman. Hardware-constrained hybrid coding of video imagery. IEEE Transactions on Aerospace and Electronic Systems, (1):71–84, 1983

    1983

  6. [6]

    pixelsplat: 3d gaussian splats from image pairs for scalable generalizable 3d reconstruction

    David Charatan, Sizhe Lester Li, Andrea Tagliasacchi, and Vincent Sitzmann. pixelsplat: 3d gaussian splats from image pairs for scalable generalizable 3d reconstruction. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 19457–19467, 2024

    2024

  7. [7]

    Mvsnerf: Fastgeneralizableradiancefieldreconstructionfrommulti-viewstereo

    AnpeiChen,ZexiangXu,FuqiangZhao,XiaoshuaiZhang,FanboXiang,JingyiYu,andHaoSu. Mvsnerf: Fastgeneralizableradiancefieldreconstructionfrommulti-viewstereo. InProceedings of the IEEE/CVF International Conference on Computer Vision, pages 14124–14133, 2021

    2021

  8. [8]

    Tensorf: Tensorial radiance fields

    Anpei Chen, Zexiang Xu, Andreas Geiger, Jingyi Yu, and Hao Su. Tensorf: Tensorial radiance fields. InEuropean conference on computer vision, pages 333–350, 2022

    2022

Show all 98 references
  1. [9]

    Photographic image synthesis with cascaded refinement networks

    Qifeng Chen and Vladlen Koltun. Photographic image synthesis with cascaded refinement networks. InProceedings of the IEEE international conference on computer vision, pages 1511–1520, 2017

    2017

  2. [10]

    Training deep nets with sublinear memory cost.arXiv preprint arXiv:1604.06174, 2016

    Tianqi Chen, Bing Xu, Chiyuan Zhang, and Carlos Guestrin. Training deep nets with sublinear memory cost.arXiv preprint arXiv:1604.06174, 2016

    2016 arXiv

  3. [11]

    One-minute video generation with test-time training

    Karan Dalal, Daniel Koceja, Gashon Hussein, Jiarui Xu, Yue Zhao, Youjin Song, Shihao Han, Ka Chun Cheung, Jan Kautz, Carlos Guestrin, et al. One-minute video generation with test-time training. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2025

    2025

  4. [12]

    Flashattention-2: Faster attention with better parallelism and work partitioning

    Tri Dao. Flashattention-2: Faster attention with better parallelism and work partitioning. InThe Twelfth International Conference on Learning Representations, 2023. 12

    2023

  5. [13]

    Objaverse: A universe of annotated 3d objects

    Matt Deitke, Dustin Schwenk, Jordi Salvador, Luca Weihs, Oscar Michel, Eli VanderBilt, Ludwig Schmidt, Kiana Ehsani, Aniruddha Kembhavi, and Ali Farhadi. Objaverse: A universe of annotated 3d objects. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Reco...

    2023

  6. [14]

    Objaverse-xl: A universe of 10m+ 3d objects

    Matt Deitke, Ruoshi Liu, Matthew Wallingford, Huong Ngo, Oscar Michel, Aditya Kusupati, Alan Fan, Christian Laforte, Vikram Voleti, Samir Yitzhak Gadre, et al. Objaverse-xl: A universe of 10m+ 3d objects. InConference on Neural Information Processing Systems, 2024

    2024

  7. [15]

    Depth-supervised nerf: Fewer views and faster training for free

    Kangle Deng, Andrew Liu, Jun-Yan Zhu, and Deva Ramanan. Depth-supervised nerf: Fewer views and faster training for free. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 12882–12891, 2022

    2022

  8. [16]

    An image is worth 16x16 words: Transformers for image recognition at scale

    Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Sylvain Gelly, et al. An image is worth 16x16 words: Transformers for image recognition at scale. InInternational Conf...

    2020

  9. [17]

    Google scanned objects: A high-quality dataset of 3d scanned household items

    LauraDowns,AnthonyFrancis,NateKoenig,BrandonKinman,RyanHickman,KristaReymann, Thomas B McHugh, and Vincent Vanhoucke. Google scanned objects: A high-quality dataset of 3d scanned household items. In2022 International Conference on Robotics and Automation (ICRA), pages 2553–256...

    2022

  10. [18]

    Deformgs: Scene flow in highly deformable scenes for deformable object manipulation

    Bardienus P Duisterhof, Zhao Mandi, Yunchao Yao, Jia-Wei Liu, Jenny Seidenschwarz, Mike Zheng Shou, Deva Ramanan, Shuran Song, Stan Birchfield, Bowen Wen, et al. Deformgs: Scene flow in highly deformable scenes for deformable object manipulation. InThe 16th International Works...

    2024

  11. [19]

    St4rtrack: Simultaneous 4d reconstruction and tracking in the world.arXiv preprint arXiv:2504.13152, 2025

    Haiwen Feng, Junyi Zhang, Qianqian Wang, Yufei Ye, Pengcheng Yu, Michael J Black, Trevor Darrell, and Angjoo Kanazawa. St4rtrack: Simultaneous 4d reconstruction and tracking in the world.arXiv preprint arXiv:2504.13152, 2025

    2025 arXiv

  12. [20]

    K-planes: Explicit radiance fields in space, time, and appearance

    Sara Fridovich-Keil, Giacomo Meanti, Frederik Rahbæk Warburg, Benjamin Recht, and Angjoo Kanazawa. K-planes: Explicit radiance fields in space, time, and appearance. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 12479–12488, 2023

    2023

  13. [21]

    Accurate, dense, and robust multiview stereopsis.IEEE transactions on pattern analysis and machine intelligence, 32(8):1362–1376, 2009

    Yasutaka Furukawa and Jean Ponce. Accurate, dense, and robust multiview stereopsis.IEEE transactions on pattern analysis and machine intelligence, 32(8):1362–1376, 2009

    2009

  14. [22]

    Gaussianflow: Splatting gaussian dynamicsfor4dcontent creation.arXiv preprint arXiv:2403.12365, 2024

    Quankai Gao, Qiangeng Xu, Zhe Cao, Ben Mildenhall, Wenchao Ma, Le Chen, Danhang Tang, andUlrich Neumann. Gaussianflow: Splatting gaussian dynamicsfor4dcontent creation.arXiv preprint arXiv:2403.12365, 2024

    2024 arXiv

  15. [23]

    Deep residual learning for image recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. InProceedings of the IEEE conference on computer vision and pattern recognition, pages 770–778, 2016

    2016

  16. [24]

    Lrm: Large reconstruction model for single image to 3d

    Yicong Hong, Kai Zhang, Jiuxiang Gu, Sai Bi, Yang Zhou, Difan Liu, Feng Liu, Kalyan Sunkavalli, Trung Bui, and Hao Tan. Lrm: Large reconstruction model for single image to 3d. InThe Twelfth International Conference on Learning Representations, 2024

    2024

  17. [25]

    2d gaussian splatting for geometrically accurate radiance fields

    Binbin Huang, Zehao Yu, Anpei Chen, Andreas Geiger, and Shenghua Gao. 2d gaussian splatting for geometrically accurate radiance fields. InSpecial Interest Group on Computer Graphics and Interactive Techniques Conference, pages 1–11, 2024

    2024

  18. [26]

    Real3d: Scaling up large reconstruction models with real-world images.arXiv preprint arXiv:2406.08479, 2024

    Hanwen Jiang, Qixing Huang, and Georgios Pavlakos. Real3d: Scaling up large reconstruction models with real-world images.arXiv preprint arXiv:2406.08479, 2024

    2024 arXiv

  19. [27]

    Rayzer: A self-supervised large view synthesis model

    Hanwen Jiang, Hao Tan, Peng Wang, Haian Jin, Yue Zhao, Sai Bi, Kai Zhang, Fujun Luan, Kalyan Sunkavalli, Qixing Huang, et al. Rayzer: A self-supervised large view synthesis model. arXiv preprint arXiv:2505.00702, 2025. 13

    2025 arXiv

  20. [28]

    Consistent4d: Consistent 360° dynamic object generation from monocular video

    Yanqin Jiang, Li Zhang, Jin Gao, Weiming Hu, and Yao Yao. Consistent4d: Consistent 360° dynamic object generation from monocular video. InThe Twelfth International Conference on Learning Representations, 2024

    2024

  21. [29]

    Lvsm: A large view synthesis model with minimal 3d inductive bias

    Haian Jin, Hanwen Jiang, Hao Tan, Kai Zhang, Sai Bi, Tianyuan Zhang, Fujun Luan, Noah Snavely, and Zexiang Xu. Lvsm: A large view synthesis model with minimal 3d inductive bias. arXiv preprint arXiv:2410.17242, 2024

    2024 arXiv

  22. [30]

    Geonerf: Generalizing nerf with geometry priors

    Mohammad Mahdi Johari, Yann Lepoittevin, and François Fleuret. Geonerf: Generalizing nerf with geometry priors. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 18365–18375, 2022

    2022

  23. [31]

    3d gaussian splatting for real-time radiance field rendering.ACM Transactions on Graphics, 42(4):1–14, 2023

    Bernhard Kerbl, Georgios Kopanas, Thomas Leimkühler, and George Drettakis. 3d gaussian splatting for real-time radiance field rendering.ACM Transactions on Graphics, 42(4):1–14, 2023

    2023

  24. [32]

    Robotseerobotdo: Imitatingarticulatedobjectmanipulationwithmonocular 4d reconstruction

    Justin Kerr, Chung Min Kim, Mingxuan Wu, Brent Yi, Qianqian Wang, Ken Goldberg, and AngjooKanazawa. Robotseerobotdo: Imitatingarticulatedobjectmanipulationwithmonocular 4d reconstruction. In8th Annual Conference on Robot Learning, 2024

    2024

  25. [33]

    Dynmf: Neural motion factorization for real-time dynamic view synthesis with 3d gaussian splatting

    Agelos Kratimenos, Jiahui Lei, and Kostas Daniilidis. Dynmf: Neural motion factorization for real-time dynamic view synthesis with 3d gaussian splatting. InEuropean Conference on Computer Vision, pages 252–269, 2024

    2024

  26. [34]

    xformers: A modular and hackable transformer modelling library

    Benjamin Lefaudeux, Francisco Massa, Diana Liskovich, Wenhan Xiong, Vittorio Caggiano, Sean Naren, Min Xu, Jieru Hu, Marta Tintore, Susan Zhang, Patrick Labatut, Daniel Haziza, Luca Wehrstedt, Jeremy Reizenstein, and Grigory Sizov. xformers: A modular and hackable transformer ...

    2022

  27. [35]

    Grounding image matching in 3d with mast3r

    Vincent Leroy, Yohann Cabon, and Jérôme Revaud. Grounding image matching in 3d with mast3r. InEuropean Conference on Computer Vision, pages 71–91, 2024

    2024

  28. [36]

    Vivid-zoo: Multi-view video generation with diffusion model

    Bing Li, Cheng Zheng, Wenxuan Zhu, Jinjie Mai, Biao Zhang, Peter Wonka, and Bernard Ghanem. Vivid-zoo: Multi-view video generation with diffusion model. InConference on Neural Information Processing Systems, pages 62189–62222, 2024

    2024

  29. [37]

    Instant3d: Fast text-to-3d with sparse-view generationandlargereconstructionmodel

    Jiahao Li, Hao Tan, Kai Zhang, Zexiang Xu, Fujun Luan, Yinghao Xu, Yicong Hong, Kalyan Sunkavalli, Greg Shakhnarovich, and Sai Bi. Instant3d: Fast text-to-3d with sparse-view generationandlargereconstructionmodel. InThe Twelfth International Conference on Learning Representati...

    2024

  30. [38]

    Diffusion4d: Fast spatial-temporal consistent 4d generationviavideodiffusionmodels

    Hanwen Liang, Yuyang Yin, Dejia Xu, Hanxue Liang, Zhangyang Wang, Konstantinos N Plataniotis, Yao Zhao, and Yunchao Wei. Diffusion4d: Fast spatial-temporal consistent 4d generationviavideodiffusionmodels. InConference on Neural Information Processing Systems, 2024

    2024

  31. [39]

    Feed-forward bullet-time reconstruction of dynamic scenes from monocular videos.arXiv preprint arXiv:2412.03526, 2024

    Hanxue Liang, Jiawei Ren, Ashkan Mirzaei, Antonio Torralba, Ziwei Liu, Igor Gilitschen- ski, Sanja Fidler, Cengiz Oztireli, Huan Ling, Zan Gojcic, et al. Feed-forward bullet-time reconstruction of dynamic scenes from monocular videos.arXiv preprint arXiv:2412.03526, 2024

    2024

  32. [40]

    Gaufre: Gaussian deformation fields for real-time dynamic novel view synthesis

    Yiqing Liang, Numair Khan, Zhengqin Li, Thu Nguyen-Phuoc, Douglas Lanman, James Tompkin, and Lei Xiao. Gaufre: Gaussian deformation fields for real-time dynamic novel view synthesis. In2025 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), pages 2642–2652. ...

    2025

  33. [41]

    Sparseneus: Fast generalizable neural surface reconstruction from sparse views

    Xiaoxiao Long, Cheng Lin, Peng Wang, Taku Komura, and Wenping Wang. Sparseneus: Fast generalizable neural surface reconstruction from sparse views. InEuropean Conference on Computer Vision, pages 210–227, 2022

    2022

  34. [42]

    InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 17724–17734, 2023

    BaoruiMa,JunshengZhou,Yu-ShenLiu,andZhizhongHan.Towardsbettergradientconsistency for neural signed distance functions via level set alignment. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 17724–17734, 2023. 14

    2023

  35. [43]

    Mixed precision training

    PauliusMicikevicius,SharanNarang,JonahAlben,GregoryDiamos,ErichElsen,DavidGarcia, Boris Ginsburg, Michael Houston, Oleksii Kuchaiev, Ganesh Venkatesh, et al. Mixed precision training. InInternational Conference on Learning Representations, 2018

    2018

  36. [44]

    Srinivasan, Matthew Tancik, Jonathan T

    Ben Mildenhall, Pratul P. Srinivasan, Matthew Tancik, Jonathan T. Barron, Ravi Ramamoorthi, andRenNg. Nerf: Representingscenesasneuralradiancefieldsforviewsynthesis. InEuropean Conference on Computer Vision, 2020

    2020

  37. [45]

    Efficient4d: Fastdynamic3dobjectgeneration from a single-view video.arXiv preprint arXiv 2401.08742, 2024

    ZijiePan,ZeyuYang,XiatianZhu,andLiZhang. Efficient4d: Fastdynamic3dobjectgeneration from a single-view video.arXiv preprint arXiv 2401.08742, 2024

    2024

  38. [46]

    Deepsdf: Learningcontinuoussigneddistancefunctionsforshaperepresentation

    Jeong Joon Park, Peter Florence, Julian Straub, Richard Newcombe, and Steven Lovegrove. Deepsdf: Learningcontinuoussigneddistancefunctionsforshaperepresentation. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 165–174, 2019

    2019

  39. [47]

    Nerfies: Deformable neural radiance fields

    Keunhong Park, Utkarsh Sinha, Jonathan T Barron, Sofien Bouaziz, Dan B Goldman, Steven M Seitz, and Ricardo Martin-Brualla. Nerfies: Deformable neural radiance fields. InProceedings of the IEEE/CVF international conference on computer vision, pages 5865–5874, 2021

    2021

  40. [48]

    Julius Plücker. Xvii. on a new geometry of space.Philosophical Transactions of the Royal Society of London, (155):725–791, 1865

  41. [49]

    Visual modeling with a hand-held camera.International Journal of Computer Vision, 59:207–232, 2004

    Marc Pollefeys, Luc Van Gool, Maarten Vergauwen, Frank Verbiest, Kurt Cornelis, Jan Tops, and Reinhard Koch. Visual modeling with a hand-held camera.International Journal of Computer Vision, 59:207–232, 2004

    2004

  42. [50]

    Detailed real-time urban 3d reconstruction from video.International Journal of Computer Vision, 78:143–167, 2008

    Marc Pollefeys, David Nistér, J-M Frahm, Amir Akbarzadeh, Philippos Mordohai, Brian Clipp, Chris Engels, David Gallup, S-J Kim, Paul Merrell, et al. Detailed real-time urban 3d reconstruction from video.International Journal of Computer Vision, 78:143–167, 2008

    2008

  43. [51]

    D-nerf: Neural radiance fields for dynamic scenes

    Albert Pumarola, Enric Corona, Gerard Pons-Moll, and Francesc Moreno-Noguer. D-nerf: Neural radiance fields for dynamic scenes. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 10318–10327, 2021

    2021

  44. [52]

    Dreamgaussian4d: Generative 4d gaussian splatting.arXiv preprint arXiv:2312.17142, 2023

    Jiawei Ren, Liang Pan, Jiaxiang Tang, Chi Zhang, Ang Cao, Gang Zeng, and Ziwei Liu. Dreamgaussian4d: Generative 4d gaussian splatting.arXiv preprint arXiv:2312.17142, 2023

    2023 arXiv

  45. [53]

    L4gm: Large 4d gaussian reconstruction model

    Jiawei Ren, Cheng Xie, Ashkan Mirzaei, Karsten Kreis, Ziwei Liu, Antonio Torralba, Sanja Fidler, Seung Wook Kim, Huan Ling, et al. L4gm: Large 4d gaussian reconstruction model. In Conference on Neural Information Processing Systems, pages 56828–56858, 2024

    2024

  46. [54]

    Structure-from-motion revisited

    Johannes L Schonberger and Jan-Michael Frahm. Structure-from-motion revisited. InProceed- ings of the IEEE conference on computer vision and pattern recognition, pages 4104–4113, 2016

    2016

  47. [55]

    Pixelwise view selection for unstructured multi-view stereo

    Johannes L Schönberger, Enliang Zheng, Jan-Michael Frahm, and Marc Pollefeys. Pixelwise view selection for unstructured multi-view stereo. InEuropean Conference on Computer Vision, pages 501–518, 2016

    2016

  48. [56]

    Maximilian Seitzer, Sjoerd van Steenkiste, Thomas Kipf, Klaus Greff, and Mehdi S. M. Sajjadi. DyST: Towards dynamic neural scene representations on real-world videos. InThe Twelfth International Conference on Learning Representations, 2024

    2024

  49. [57]

    Phototourism: exploringphotocollections in 3d.ACM Transactions on Graphics (TOG), 25(3):835–846, 2006

    NoahSnavely,StevenMSeitz,andRichardSzeliski. Phototourism: exploringphotocollections in 3d.ACM Transactions on Graphics (TOG), 25(3):835–846, 2006

    2006

  50. [58]

    Generalizable patch-based neural rendering

    Mohammed Suhail, Carlos Esteves, Leonid Sigal, and Ameesh Makadia. Generalizable patch-based neural rendering. InEuropean Conference on Computer Vision, pages 156–174, 2022

    2022

  51. [59]

    Light field neural rendering

    Mohammed Suhail, Carlos Esteves, Leonid Sigal, and Ameesh Makadia. Light field neural rendering. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 8269–8279, 2022. 15

    2022

  52. [60]

    Lgm: Large multi-view gaussian model for high-resolution 3d content creation

    Jiaxiang Tang, Zhaoxi Chen, Xiaokang Chen, Tengfei Wang, Gang Zeng, and Ziwei Liu. Lgm: Large multi-view gaussian model for high-resolution 3d content creation. InEuropean Conference on Computer Vision, pages 1–18, 2024

    2024

  53. [61]

    Mv-dust3r+: Single-stage scene reconstruction from sparse views in 2 seconds

    ZhenggangTang,YuchenFan,DilinWang,HongyuXu,RakeshRanjan,AlexanderSchwing,and Zhicheng Yan. Mv-dust3r+: Single-stage scene reconstruction from sparse views in 2 seconds. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024

    2024

  54. [62]

    Mononerf: Learning a generalizable dynamic radiancefieldfrommonocularvideos

    Fengrui Tian, Shaoyi Du, and Yueqi Duan. Mononerf: Learning a generalizable dynamic radiancefieldfrommonocularvideos. InProceedings of the IEEE/CVF International Conference on Computer Vision, pages 17903–17913, 2023

    2023

  55. [63]

    Non-rigid neural radiance fields: Reconstruction and novel view synthesis of a dynamic scene from monocular video

    Edgar Tretschk, Ayush Tewari, Vladislav Golyanik, Michael Zollhöfer, Christoph Lassner, and Christian Theobalt. Non-rigid neural radiance fields: Reconstruction and novel view synthesis of a dynamic scene from monocular video. InProceedings of the IEEE/CVF International Confer...

    2021

  56. [64]

    Attention is all you need

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. Attention is all you need. InConference on Neural Information Processing Systems, 2017

    2017

  57. [65]

    Sv3d: Novel multi-view synthesis and 3d generation from a single image using latent video diffusion

    Vikram Voleti, Chun-Han Yao, Mark Boss, Adam Letts, David Pankratz, Dmitry Tochilkin, Christian Laforte, Robin Rombach, and Varun Jampani. Sv3d: Novel multi-view synthesis and 3d generation from a single image using latent video diffusion. InEuropean Conference on Computer Vis...

    2024

  58. [66]

    Vggt: Visual geometry grounded transformer

    JianyuanWang,MinghaoChen,NikitaKaraev,AndreaVedaldi,ChristianRupprecht,andDavid Novotny. Vggt: Visual geometry grounded transformer. InProceedings of the Computer Vision and Pattern Recognition Conference, pages 5294–5306, 2025

    2025

  59. [67]

    Pf-lrm: Pose-free large reconstruction model for joint pose and shape prediction

    Peng Wang, Hao Tan, Sai Bi, Yinghao Xu, Fujun Luan, Kalyan Sunkavalli, Wenping Wang, Zexiang Xu, and Kai Zhang. Pf-lrm: Pose-free large reconstruction model for joint pose and shape prediction. InThe Twelfth International Conference on Learning Representations, 2024

    2024

  60. [68]

    Ibrnet: Learning multi-view image-based rendering

    Qianqian Wang, Zhicheng Wang, Kyle Genova, Pratul P Srinivasan, Howard Zhou, Jonathan T Barron, Ricardo Martin-Brualla, Noah Snavely, and Thomas Funkhouser. Ibrnet: Learning multi-view image-based rendering. InProceedings of the IEEE/CVF Conference on Computer Vision and Patte...

    2021

  61. [69]

    Shape of motion: 4d reconstruction from a single video.arXiv preprint arXiv:2407.13764, 2024

    Qianqian Wang, Vickie Ye, Hang Gao, Jake Austin, Zhengqi Li, and Angjoo Kanazawa. Shape of motion: 4d reconstruction from a single video.arXiv preprint arXiv:2407.13764, 2024

    2024

  62. [70]

    Continuous 3d perception model with persistent state

    Qianqian Wang, Yifei Zhang, Aleksander Holynski, Alexei A Efros, and Angjoo Kanazawa. Continuous 3d perception model with persistent state. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2025

    2025

  63. [71]

    Dust3r: Geometric 3d vision made easy

    Shuzhe Wang, Vincent Leroy, Yohann Cabon, Boris Chidlovskii, and Jerome Revaud. Dust3r: Geometric 3d vision made easy. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 20697–20709, 2024

    2024

  64. [72]

    Image quality assessment: from error visibility to structural similarity.IEEE transactions on image processing, 13(4): 600–612, 2004

    Zhou Wang, Alan C Bovik, Hamid R Sheikh, and Eero P Simoncelli. Image quality assessment: from error visibility to structural similarity.IEEE transactions on image processing, 13(4): 600–612, 2004

    2004

  65. [73]

    Controlling space and time with diffusion models

    Daniel Watson, Saurabh Saxena, Lala Li, Andrea Tagliasacchi, and David J Fleet. Controlling space and time with diffusion models. InThe Thirteenth International Conference on Learning Representations, 2024

    2024

  66. [74]

    Meshlrm: Large reconstruction model for high-quality mesh.arXiv preprint arXiv:2404.12385, 2024

    Xinyue Wei, Kai Zhang, Sai Bi, Hao Tan, Fujun Luan, Valentin Deschaintre, Kalyan Sunkavalli, Hao Su, and Zexiang Xu. Meshlrm: Large reconstruction model for high-quality mesh.arXiv preprint arXiv:2404.12385, 2024. 16

    2024 arXiv

  67. [75]

    4d gaussian splatting for real-time dynamic scene rendering

    Guanjun Wu, Taoran Yi, Jiemin Fang, Lingxi Xie, Xiaopeng Zhang, Wei Wei, Wenyu Liu, Qi Tian, and Xinggang Wang. 4d gaussian splatting for real-time dynamic scene rendering. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 20310–20320, 2024

    2024

  68. [76]

    Cat4d: Create anything in 4d with multi-view video diffusion models

    Rundi Wu, Ruiqi Gao, Ben Poole, Alex Trevithick, Changxi Zheng, Jonathan T Barron, and Aleksander Holynski. Cat4d: Create anything in 4d with multi-view video diffusion models. arXiv preprint arXiv:2411.18613, 2024

    2024 arXiv

  69. [77]

    Lrm-zero: Training large reconstruction models with synthesized data

    Desai Xie, Sai Bi, Zhixin Shu, Kai Zhang, Zexiang Xu, Yi Zhou, Soren Pirk, Arie Kaufman, Xin Sun, and Hao Tan. Lrm-zero: Training large reconstruction models with synthesized data. InThe Thirty-eighth Annual Conference on Neural Information Processing Systems, 2024

    2024

  70. [78]

    Sv4d: Dy- namic 3d content generation with multi-frame and multi-view consistency.arXiv preprint arXiv:2407.17470, 2024

    Yiming Xie, Chun-Han Yao, Vikram Voleti, Huaizu Jiang, and Varun Jampani. Sv4d: Dy- namic 3d content generation with multi-frame and multi-view consistency.arXiv preprint arXiv:2407.17470, 2024

    2024 arXiv

  71. [79]

    Dmv3d: Denoising multi-view diffusion using 3d large reconstruction model

    Yinghao Xu, Hao Tan, Fujun Luan, Sai Bi, Peng Wang, Jiahao Li, Zifan Shi, Kalyan Sunkavalli, Gordon Wetzstein, Zexiang Xu, et al. Dmv3d: Denoising multi-view diffusion using 3d large reconstruction model. InThe Twelfth International Conference on Learning Representations, 2024

    2024

  72. [80]

    Fast3r: Towards 3d reconstruction of 1000+ images in one forward pass

    Jianing Yang, Alexander Sax, Kevin J Liang, Mikael Henaff, Hao Tang, Ang Cao, Joyce Chai, Franziska Meier, and Matt Feiszli. Fast3r: Towards 3d reconstruction of 1000+ images in one forward pass. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2025

    2025

  73. [81]

    Storm: Spatio-temporal reconstruction model for large-scale outdoor scenes

    Jiawei Yang, Jiahui Huang, Yuxiao Chen, Yan Wang, Boyi Li, Yurong You, Apoorva Sharma, Maximilian Igl, Peter Karkus, Danfei Xu, et al. Storm: Spatio-temporal reconstruction model for large-scale outdoor scenes. InThe Thirteenth International Conference on Learning Representati...

    2025

  74. [82]

    Real-time photorealistic dynamic scene representationandrenderingwith4dgaussiansplatting

    Zeyu Yang, Hongye Yang, Zijie Pan, and Li Zhang. Real-time photorealistic dynamic scene representationandrenderingwith4dgaussiansplatting. InThe Twelfth International Conference on Learning Representations, 2024

    2024

  75. [83]

    Sv4d 2.0: Enhancing spatio-temporal consistency in multi-view video diffusion for high-quality 4d generation.arXiv preprint arXiv:2503.16396, 2025

    Chun-Han Yao, Yiming Xie, Vikram Voleti, Huaizu Jiang, and Varun Jampani. Sv4d 2.0: Enhancing spatio-temporal consistency in multi-view video diffusion for high-quality 4d generation.arXiv preprint arXiv:2503.16396, 2025

    2025 arXiv

  76. [84]

    Neural cages for detail-preserving 3d deformations

    Wang Yifan, Noam Aigerman, Vladimir G Kim, Siddhartha Chaudhuri, and Olga Sorkine- Hornung. Neural cages for detail-preserving 3d deformations. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 75–83, 2020

    2020

  77. [85]

    4dgen: Grounded 4d content generation with spatial-temporal consistency.arXiv preprint arXiv:2312.17225, 2023

    Yuyang Yin, Dejia Xu, Zhangyang Wang, Yao Zhao, and Yunchao Wei. 4dgen: Grounded 4d content generation with spatial-temporal consistency.arXiv preprint arXiv:2312.17225, 2023

    2023 arXiv

  78. [86]

    pixelnerf: Neural radiance fields from one or few images

    Alex Yu, Vickie Ye, Matthew Tancik, and Angjoo Kanazawa. pixelnerf: Neural radiance fields from one or few images. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 4578–4587, 2021

    2021

  79. [87]

    Mip-splatting: Alias-free 3d gaussian splatting

    Zehao Yu, Anpei Chen, Binbin Huang, Torsten Sattler, and Andreas Geiger. Mip-splatting: Alias-free 3d gaussian splatting. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 19447–19456, 2024

    2024

  80. [88]

    Stag4d: Spatial-temporal anchored generative 4d gaussians

    Yifei Zeng, Yanqin Jiang, Siyu Zhu, Yuanxun Lu, Youtian Lin, Hao Zhu, Weiming Hu, Xun Cao, and Yao Yao. Stag4d: Spatial-temporal anchored generative 4d gaussians. InEuropean Conference on Computer Vision, pages 163–179, 2024

    2024

  81. [89]

    4diffusion: Multi-view video diffusion model for 4d generation

    HaiyuZhang,XinyuanChen,YaohuiWang,XihuiLiu,YunhongWang,andYuQiao. 4diffusion: Multi-view video diffusion model for 4d generation. InNeural Information Processing Systems, volume 37, pages 15272–15295, 2024. 17

    2024

  82. [90]

    Monst3r: Asimpleapproachforestimatinggeometryinthe presence of motion

    Junyi Zhang, Charles Herrmann, Junhwa Hur, Varun Jampani, Trevor Darrell, Forrester Cole, DeqingSun,andMing-HsuanYang. Monst3r: Asimpleapproachforestimatinggeometryinthe presence of motion. InThe Thirteenth International Conference on Learning Representations, 2025

    2025

  83. [91]

    Arf: Artistic radiance fields

    Kai Zhang, Nick Kolkin, Sai Bi, Fujun Luan, Zexiang Xu, Eli Shechtman, and Noah Snavely. Arf: Artistic radiance fields. InEuropean Conference on Computer Vision, pages 717–733, 2022

    2022

  84. [92]

    Gs-lrm: Large reconstruction model for 3d gaussian splatting

    Kai Zhang, Sai Bi, Hao Tan, Yuanbo Xiangli, Nanxuan Zhao, Kalyan Sunkavalli, and Zexiang Xu. Gs-lrm: Large reconstruction model for 3d gaussian splatting. InEuropean Conference on Computer Vision, pages 1–19, 2024

    2024

  85. [93]

    Theunreasonable effectiveness of deep features as a perceptual metric

    RichardZhang,PhillipIsola,AlexeiAEfros,EliShechtman,andOliverWang. Theunreasonable effectiveness of deep features as a perceptual metric. InProceedings of the IEEE conference on computer vision and pattern recognition, pages 586–595, 2018

    2018

  86. [94]

    Test-time training done right.arXiv preprint arXiv:2505.23884, 2025

    Tianyuan Zhang, Sai Bi, Yicong Hong, Kai Zhang, Fujun Luan, Songlin Yang, Kalyan Sunkavalli, William T Freeman, and Hao Tan. Test-time training done right.arXiv preprint arXiv:2505.23884, 2025

    2025 arXiv

  87. [95]

    Susskind, and Alex Schwing

    Xiaoming Zhao, R Alex Colburn, Fangchang Ma, Miguel Ángel Bautista, Joshua M. Susskind, and Alex Schwing. Pseudo-generalized dynamic view synthesis from a video. InThe Twelfth International Conference on Learning Representations, 2024

    2024

  88. [96]

    Tesseract: Learning 4d embodied world models.arXiv preprint arXiv:2504.20995, 2025

    Haoyu Zhen, Qiao Sun, Hongxin Zhang, Junyan Li, Siyuan Zhou, Yilun Du, and Chuang Gan. Tesseract: Learning 4d embodied world models.arXiv preprint arXiv:2504.20995, 2025

    2025 arXiv

  89. [97]

    Drivable 3d gaussian avatars

    Wojciech Zielonka, Timur Bagautdinov, Shunsuke Saito, Michael Zollhöfer, Justus Thies, and Javier Romero. Drivable 3d gaussian avatars. InInternational Conference on 3D Vision, 2025

    2025

  90. [98]

    Long-lrm: Long-sequence large reconstruction model for wide-coverage gaussian splats.arXiv preprint arXiv:2410.12781, 2024

    Chen Ziwen, Hao Tan, Kai Zhang, Sai Bi, Fujun Luan, Yicong Hong, Li Fuxin, and Zexiang Xu. Long-lrm: Long-sequence large reconstruction model for wide-coverage gaussian splats.arXiv preprint arXiv:2410.12781, 2024. 18 A 4D-LRM Implementation and Training Details A.1 4DGS Param...

    2024 arXiv

Pith tools

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