Pith. sign in

REVIEW 4 major objections 5 minor 1 cited by

OmniSplat: Taming Feed-Forward 3D Gaussian Splatting for Omnidirectional Images with Editable Capabilities

T0 review · 4 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read OmniSplat claims that reshaping omnidirectional images onto a Yin-Yang grid lets an off-the-shelf perspective-trained feed-forward 3D Gaussian splatting network reconstruct a full 360-degree scene from two photos, with no fine-tuning.

desk verdict A genuinely new training-free adaptation of perspective feed-forward 3DGS to 360 images via Yin-Yang decomposition; the result is plausible but the manual reference-view selection and circular segmentation evaluation keep it from being fully convincing. read the letter →

arxiv 2412.16604 v2 pith:OVQXZLRO submitted 2024-12-21 cs.CV

Suyoung Lee , Jaeyoung Chung , Kihoon Kim , Jaeyoo Huh , Gunhee Lee , Minsoo Lee , Kyoung Mu Lee This is my paper · ORCID
classification cs.CV
keywords omnidirectionalimages3DGaussiansplattingfeed-forwardreconstructionYin-Yanggridnovelviewsynthesisequirectangularprojectionsparse-viewattention-basedsegmentation
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

OmniSplat claims that an unmodified perspective-trained feed-forward 3D Gaussian splatting network can reconstruct a complete 360-degree scene from just two omnidirectional photos, as long as the images are first reshaped onto a Yin-Yang grid. The grid's quasi-uniform square tiles look enough like perspective images that the pretrained model's priors transfer, eliminating the need for omnidirectional training data or fine-tuning. The paper also introduces a Yin-Yang rasterizer that renders the two tiles separately and stitches them in pixel space, removing near-pole stripe artifacts. If the claim holds, 360-degree cameras become drop-in inputs for instant sparse-view 3D reconstruction, and the attention maps used for depth estimation double as a free multiview-consistent segmentation signal for 3D editing.

What carries the argument

The load-bearing object is the Yin-Yang grid, an overset spherical coordinate system made of two overlapping square lattices: Yin, defined by $\theta\in[-\pi/4,\pi/4]$ and $\phi\in[-3\pi/4,3\pi/4]$, and Yang, obtained by rotating Yin with the matrix $M$. Its two properties do the work: quasi-uniformness makes each tile nearly perspective-like, so a CNN or Transformer trained on perspective photos can be reused unchanged, while the structured square lattice keeps the grid CNN-compatible, unlike icosahedral or triangular tilings. The companion mechanism is the Yin-Yang rasterizer, which renders the two tiles separately from the predicted 3D Gaussians, normalizes each by its $\alpha$ map, and warps and combines them into one equirectangular image, suppressing the stripe artifacts caused by non-uniform sampling near the poles.

What would settle it

Evaluate on the same six benchmarks using automatically selected or randomly chosen reference pairs instead of the manually selected low-occlusion pairs, and compare PSNR against the cubemap and equirectangular baselines; if the margin shrinks to insignificance or reverses, the training-free advantage would be refuted.

Watch

Extended reading notes

Core claim

The paper's central claim is that feed-forward 3D Gaussian splatting from omnidirectional images can be achieved without any training or fine-tuning by changing the image grid rather than the network. Each equirectangular input is cut into two quasi-uniform square tiles, the Yin grid and the Yang grid (the latter is the former rotated by the matrix $M$), so the four tiles from two reference views look enough like perspective photographs that a pretrained perspective-image estimator extracts correct features without adaptation. A mask-weighted cross-view attention warps each grid feature to the other view along depth candidates and builds cost volumes for all four grid combinations, preventing information loss at grid boundaries. The estimated Gaussians are then rendered not onto an equirectangular raster but separately onto Yin and Yang images, which are divided by their alpha maps and fused in pixel space into the final omnidirectional view. On six datasets the paper reports that this pipeline attains the best PSNR-runtime trade-off among feed-forward methods, and that a short 100-step color/opacity optimization makes the result competitive with or better than a 30,000-step optimization baseline; the same attention correspondences also yield multiview-consistent segmentation used for clean 3D editing.

Load-bearing premise

The load-bearing premise is that the hand-picked reference pairs, chosen to minimize occluded regions, represent how the method would actually be used, and that the reported gains survive ordinary or automatically chosen view pairs.

Editorial extensions

If this is right

  • Two reference omnidirectional images synthesize novel 360-degree views in roughly 0.5 seconds, fast enough for interactive capture and preview.
  • The method beats cubemap- and equirectangular-adapted perspective feed-forward baselines on PSNR in most of the six tested datasets, while running substantially faster than the perspective baselines that must process many cubemap faces.
  • A short 100-step optimization that updates only color and opacity (about 12 seconds) raises quality further without the overfitting that plagues optimization-only omnidirectional 3DGS.
  • Attention-score correspondence gives multiview-consistent segmentation that supports clean removal and editing of Gaussians with no extra computation.
  • The same decomposition-and-rasterization recipe transfers to large-motion capture settings (360Roam, OmniScenes, 360VO) as well as dense small-baseline video datasets.

Reading between the lines

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

  • If the wrapper is as generic as it looks, any perspective-trained 3D estimator could be ported to 360-degree input by the same Yin-Yang front-end, not just the specific feed-forward 3DGS network tested here.
  • The hand-picked reference pairs in the evaluation leave open whether the margin survives automatic or arbitrary view selection; testing random and occluded pairs would settle how much of the gain comes from the grid versus the favorable baseline setup.
  • The acknowledged horizontal interpolation when converting a Yang tile back to equirectangular space is the most likely quality bottleneck, so seam-aware blending or an invertible grid parameterization is a natural next step.
  • Segmentation from attention scores could grow into a full zero-shot 3D editing pipeline, where selecting an object in one view edits or removes it consistently across all views without per-scene optimization; the paper shows removal but not yet an end-to-end editing interface.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. OmniSplat proposes a training-free feed-forward 3D Gaussian splatting pipeline for omnidirectional novel-view synthesis. It decomposes each equirectangular input into two Yin-Yang grid images, feeds the four decomposed images (two views times two grids) into an unmodified MVSplat network with a modified cross-view attention that mixes features across grids under a spherical sweep, and renders novel views with a Yin-Yang rasterizer. The method is evaluated on six omnidirectional datasets against perspective-adapted and omnidirectional-adapted feed-forward baselines as well as the optimization-based ODGS, and a secondary attention-based segmentation/editing capability is presented.

Significance. If the central claim holds, this is a useful and practical contribution: it shows that a perspective-trained feed-forward 3DGS model can be applied to omnidirectional images without fine-tuning, using a principled spherical representation (Yin-Yang grid) and a matching rasterizer. The paper is commendably transparent in several ways: it provides per-scene supplementary tables, an ablation separating the encoder and rasterizer contributions (Table 2), and the method itself is not circular in the novel-view-synthesis setting because it reuses pretrained MVSplat weights and a published omnidirectional rasterizer without fitting to the target views. The main weaknesses are in the evaluation protocol: the reference views are manually selected to favor low occlusion, no error bars or significance tests are given, and the segmentation evaluation is circular. These issues do not invalidate the technical core, but they currently prevent the strongest claims from being fully supported.

major comments (4)
  1. [Section 4, Experimental details; Tables B-G] The central claim that OmniSplat 'demonstrates higher reconstruction accuracy than existing feed-forward networks trained on perspective images' is evaluated only under manually selected reference pairs. The text states that 'we select two indices as reference views for each scene where the two images are taken from a distance to minimize the portion of occluded regions,' and the supplementary tables confirm hand-picked indices such as OmniBlender frames 14 and 86, Ricoh360 frames 35 and 69, and OmniPhotos frames 0 and 35. This protocol does not reflect automatic or typical capture, where reference views may be adjacent or otherwise not chosen to minimize occlusion. I request an evaluation with fixed, random, or automatically selected reference pairs, reporting the distribution of results (mean, standard deviation, or paired per-scene differences) and including adjacent-frame pairs. Without this, the reported gains over PixelSplat(P) and MVSplat(P) may be contingent on the favorable wide-baseline, low-occlusion choice.
  2. [Table 1 and Tables B-G] No error bars or statistical significance tests are reported. The per-scene tables show that the aggregate margins are small and inconsistent: on OmniPhotos, OmniSplat is 1.35 dB below PixelSplat(P) (18.59 vs 19.94), and on OmniBlender, PixelSplat(P) beats OmniSplat in several scenes (e.g., bistro square 20.11 vs 18.43, restroom 24.95 vs 23.02). A paired test or a per-scene visualization of differences is needed to support the claim that OmniSplat is broadly 'higher reconstruction accuracy' rather than better on a selected subset.
  3. [Section 4.2, Table 3] The segmentation evaluation is circular. The paper states: 'Due to the absence of ground truth segment maps in the datasets, we applied DEVA tracking over full image sequences and treated the result as pseudo ground truth for evaluation.' DEVA is then used as the baseline in Table 3, so the comparison measures how well OmniSplat matches DEVA-derived pseudo labels against DEVA itself. This cannot support the claim of 'superior segmentation correspondence.' Please evaluate on human-annotated segment correspondences, or use an independent matching oracle, or clearly reframe the result as agreement with DEVA rather than superiority.
  4. [Section 4.1, Table 1] The text overstates the comparison with ODGS and with the feed-forward baselines. Section 1 says OmniSplat 'shows the fastest synthesis speed while reaching the highest PSNR than any other models,' and Section 4.1 says that even after 30,000 iterations ODGS results 'do not outperform than the feed-forward networks.' Table 1 contradicts this: ODGS has higher PSNR than OmniSplat on OmniBlender (22.23 vs 21.22), OmniPhotos (20.25 vs 18.59), 360Roam (18.72 vs 17.53), and 360VO (22.66 vs 20.63); and PixelSplat(P) has higher PSNR than OmniSplat on OmniPhotos. Please revise these statements to be scoped to feed-forward comparisons or to the PSNR-runtime trade-off, and correct the ODGS discussion.
minor comments (5)
  1. [Section 3.1, Eq. (1)] The definition of the Yin grid uses both an elevation interval and a 270-degree azimuth interval; it would help to state explicitly which spherical coordinates are used (elevation vs latitude, azimuth convention) and to include a small schematic of the Yin-Yang overlap, since the geometric setup is central to the method.
  2. [Section 4.2] The attention-based segment matching is described only qualitatively ('matches the most focused pixel'). Please provide the exact algorithm: how attention scores are aggregated across depth candidates and Yin/Yang grids, how the query points are propagated, and how labels are assigned to Gaussians.
  3. [Section 4.1, Table 1] The PanoGRF baseline is not tagged with an input-domain label, and it is unclear whether it is used as released or adapted for this task. Please specify the exact protocol for this baseline.
  4. [Section 4, Experimental details] The reported runtimes should state the hardware and the exact pipeline endpoints (e.g., whether pose estimation via OpenMVG and image I/O are excluded). The runtime comparison is a major selling point, so this information is needed for reproducibility.
  5. [Section 5, Limitations] The limitation about interpolation artifacts when transforming the Yang image to the omnidirectional domain is not quantified. A short experiment or even a per-dataset LPIPS breakdown for the seam region would help readers understand how much this interpolation affects the reported numbers.

Circularity Check

1 steps flagged · score 2.0 of 10

Central feed-forward reconstruction claim is self-contained; only the secondary segmentation evaluation is circular because DEVA supplies the pseudo-ground truth used to claim superiority over DEVA.

  1. other [Section 4.2, Multiview Consistent Segmentation / Table 3]
    "Due to the absence of ground truth segment maps in the datasets, we applied DEV A tracking over full image sequences and treated the result as pseudo ground truth for evaluation."

    The same paragraph and Table 3 use this pseudo-ground truth to claim that OmniSplat 'achieves superior segmentation correspondence' over DEVA. Since the reference labels are produced by DEVA itself, the mIoU comparison to DEVA is not an independent benchmark: the target of the comparison is also the source of the ground truth. This makes the claimed superiority over DEVA circular by construction for that experiment. It does not affect the central novel-view-synthesis claim, which is evaluated against held-out ground-truth images with external baselines.

full rationale

The main derivation (Sections 3.1-3.2) is a composition of external components: pretrained MVSplat weights, the Yin-Yang coordinate transform [16], and the ODGS omnidirectional rasterizer [20]. No parameter is fitted to the novel-view test set, and the reported PSNR/SSIM/LPIPS numbers compare against held-out omnidirectional test views from six external datasets. The manual selection of wide-baseline reference views (Section 4, Experimental details) is a protocol weakness and may overstate the advantage, but it is not circular: the target images are still held out and the baselines use the same reference choices. The only genuine circularity is in the secondary segmentation experiment, where DEVA is used to generate pseudo ground truth and then OmniSplat is declared superior to DEVA. That step is circular, but it is not load-bearing for the central feed-forward 3DGS claim; hence the low score.

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

No new physical or conceptual entities are introduced; the method recombines known components (Yin-Yang grid, MVSplat, ODGS rasterizer). The main load-bearing assumptions are the domain-adaptation power of Yin-Yang grids and the generalizability of pretrained features, both supported only by empirical evaluation.

free parameters (3)
  • depth candidate range d_near, d_far = 1, 100
    Manually set in Section 4; not optimized per scene.
  • Yin-Yang grid resolution (H_n, W_n)
    Not reported; design choice affecting sampling density and results.
  • Number of depth candidates
    Inherited from MVSplat but not specified; affects cost volume quality.
assumptions (3)
  • domain assumption MVSplat pretrained on perspective images generalizes to Yin-Yang decomposed images and spherical epipolar warping.
    The entire training-free adaptation depends on this; validated only empirically on six datasets.
  • domain assumption Yin-Yang grid patches are similar enough to perspective images for the encoder to extract correct features.
    Core motivation; supported by ablation (Table 2) but not proven.
  • standard math ODGS omnidirectional rasterizer can be reused for Yin-Yang rendering by applying rotation M.
    Rotation of camera pose is a standard geometric transform.

how reviews work

0 comments
Cite this review

Pith. "Pith review of OmniSplat: Taming Feed-Forward 3D Gaussian Splatting for Omnidirectional Images with Editable Capabilities." pith.science (2026). https://pith.science/paper/OVQXZLRO

@misc{pith2026241216604,
  author       = {Pith},
  title        = {Pith review of: OmniSplat: Taming Feed-Forward 3D Gaussian Splatting for Omnidirectional Images with Editable Capabilities},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/OVQXZLRO}},
  note         = {Machine review of arXiv:2412.16604}
}
read the original abstract

Feed-forward 3D Gaussian splatting (3DGS) models have gained significant popularity due to their ability to generate scenes immediately without needing per-scene optimization. Although omnidirectional images are becoming more popular since they reduce the computation required for image stitching to composite a holistic scene, existing feed-forward models are only designed for perspective images. The unique optical properties of omnidirectional images make it difficult for feature encoders to correctly understand the context of the image and make the Gaussian non-uniform in space, which hinders the image quality synthesized from novel views. We propose OmniSplat, a training-free fast feed-forward 3DGS generation framework for omnidirectional images. We adopt a Yin-Yang grid and decompose images based on it to reduce the domain gap between omnidirectional and perspective images. The Yin-Yang grid can use the existing CNN structure as it is, but its quasi-uniform characteristic allows the decomposed image to be similar to a perspective image, so it can exploit the strong prior knowledge of the learned feed-forward network. OmniSplat demonstrates higher reconstruction accuracy than existing feed-forward networks trained on perspective images. Our project page is available on: https://robot0321.github.io/omnisplat/index.html.

Figures

Figures reproduced from arXiv: 2412.16604 by the authors.

Figure 1
Figure 1. PSNR-runtime trade-off. A chart of reconstruction PSNR-runtime trade-off in novel view image on OmniBlender [7]. OmniSplat shows the best trade-off compared to the original feed￾forward networks for perspective images. works [3, 4, 32], which are capable of generating 3D Gaus￾sian splatting (3DGS) representations directly from a few input images without scene-wise optimization [17]. These models estimate plausible 3… view at source ↗
Figure 2
Figure 2. The overall process of OmniSplat. The two reference omnidirectional images are decomposed into Yin-Yang images, and the cross-view attention is conducted across grids along with epipolar lines to compose cost volume. The 3DGS parameters are estimated and Yin-Yang images are rasterized from the novel view. The two images are combined to synthesize the final omnidirectional image. In cross-view attention, we present r… view at source ↗
Figure 3
Figure 3. Qualitative comparison. Novel view synthesized image examples in various datasets. Each scene is brought from OmniBlender, Ricoh360, and OmniPhotos, respectively. Best viewed when zoomed in. Encoder Rasterizer PSNR↑ SSIM↑ LPIPS↓ Yin-Yang Omnidirectional 13.30 0.4836 0.4965 Omnidirectional Yin-Yang 15.97 0.5717 0.4001 Yin-Yang Yin-Yang 18.59 0.6195 0.4115 (a) OmniPhotos [2] Encoder Rasterizer PSNR↑ SSIM↑ LPIPS↓ Yin-Y… view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Visualization of segment matching. We visualize the matched segment samples among the source (top) and the target (bottom) views. The stars in the image indicate query points for the user to segment objects containing the stars. Dataset OmniBlender Ricoh360 OmniPhotos …

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

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

  1. PriOr-Flow: Enhancing Primitive Panoramic Optical Flow with Orthogonal View

    cs.CV 2025-06 conditional novelty 6.0 of 10

    A dual-branch optical flow network using a 90-degree rotated 'orthogonal' view reduces polar distortion errors and sets new state-of-the-art results on MPFDataset and FlowScape.

Reference graph

Works this paper leans on

41 extracted references · 35 canonical work pages · cited by 1 Pith paper

  1. [1]

    Loopsparsegs: Loop based sparse-view friendly gaussian splatting

    Zhenyu Bao, Guibiao Liao, Kaichen Zhou, Kanglin Liu, Qing Li, and Guoping Qiu. Loopsparsegs: Loop based sparse-view friendly gaussian splatting. arXiv preprint arXiv:2408.00254,

    arXiv

  2. [2]

    Omniphotos: casual 360 vr photography

    Tobias Bertel, Mingze Yuan, Reuben Lindroos, and Christian Richardt. Omniphotos: casual 360 vr photography. ACM TOG, 2020. 5, 7, 12

    work page 2020

  3. [3]

    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. In CVPR,

    work page

  4. [4]

    Mvsplat: Efficient 3d gaussian splatting from sparse multi-view images

    Yuedong Chen, Haofei Xu, Chuanxia Zheng, Bohan Zhuang, Marc Pollefeys, Andreas Geiger, Tat-Jen Cham, and Jianfei Cai. Mvsplat: Efficient 3d gaussian splatting from sparse multi-view images. In ECCV, 2024. 1, 2, 4, 5, 6, 7, 8, 12

    work page 2024

  5. [5]

    Panogrf: Generalizable spherical radiance fields for wide-baseline panoramas

    Zheng Chen, Yan-Pei Cao, Yuan-Chen Guo, Chen Wang, Ying Shan, and Song-Hai Zhang. Panogrf: Generalizable spherical radiance fields for wide-baseline panoramas. In NeurIPS, 2023. 6

    work page 2023

  6. [6]

    Deva: Tracking anything with decoupled video segmentation

    Ho Kei Cheng, Seoung Wug Oh, Brian Price, Alexander Schwing, and Joon-Young Lee. Deva: Tracking anything with decoupled video segmentation. In ICCV, 2023. 8

    work page 2023

  7. [7]

    Bal- anced spherical grid for egocentric view synthesis

    Changwoon Choi, Sang Min Kim, and Young Min Kim. Bal- anced spherical grid for egocentric view synthesis. In CVPR,

    work page

  8. [8]

    Depth- regularized optimization for 3d gaussian splatting in few-shot images

    Jaeyoung Chung, Jeongtaek Oh, and Kyoung Mu Lee. Depth- regularized optimization for 3d gaussian splatting in few-shot images. In CVPRW, 2024. 2

    work page 2024

Show all 41 references
  1. [9]

    Gauge equivariant convolutional networks and the icosahedral cnn

    Taco Cohen, Maurice Weiler, Berkay Kicanaoglu, and Max Welling. Gauge equivariant convolutional networks and the icosahedral cnn. In ICML, 2019. 2, 3

    2019

  2. [10]

    Blender - a 3D modelling and rendering package

    Blender Online Community. Blender - a 3D modelling and rendering package. Blender Foundation, Stichting Blender Foundation, Amsterdam, 2018. 5

    2018

  3. [11]

    Spherenet: Learning spherical representations for detection and classification in omnidirectional images

    Benjamin Coors, Alexandru Paul Condurache, and Andreas Geiger. Spherenet: Learning spherical representations for detection and classification in omnidirectional images. In ECCV, 2018. 3

    2018

  4. [12]

    360vo: Visual odometry using a single 360 camera

    Huajian Huang and Sai-Kit Yeung. 360vo: Visual odometry using a single 360 camera. In ICRA, 2022. 5, 12

    2022

  5. [13]

    Real-time omnidirectional roaming in large scale indoor scenes

    Huajian Huang, Yingshu Chen, Tianjia Zhang, and Sai-Kit Yeung. Real-time omnidirectional roaming in large scale indoor scenes. In SIGGRAPH Asia, 2022. 5, 12

    2022

  6. [14]

    On the error analysis of 3d gaussian splatting and an optimal projection strategy

    Letian Huang, Jiayang Bai, Jie Guo, Yuanqi Li, and Yanwen Guo. On the error analysis of 3d gaussian splatting and an optimal projection strategy. In ECCV, 2024. 2, 5

    2024

  7. [15]

    Effective rank analysis and regularization for enhanced 3d gaussian splatting

    Junha Hyung, Susung Hong, Sungwon Hwang, Jaeseong Lee, Jaegul Choo, and Jin-Hwa Kim. Effective rank analysis and regularization for enhanced 3d gaussian splatting. InNeurIPS,

  8. [16]

    yin-yang grid

    Akira Kageyama and Tetsuya Sato. “yin-yang grid”: An overset grid in spherical geometry.Geochemistry, Geophysics, Geosystems, 5(9), 2004. 2, 3

    2004

  9. [17]

    3d gaussian splatting for real-time radiance field rendering

    Bernhard Kerbl, Georgios Kopanas, Thomas Leimk¨uhler, and George Drettakis. 3d gaussian splatting for real-time radiance field rendering. ACM TOG, 2023. 1, 2

    2023

  10. [18]

    Graph-based classifi- cation of omnidirectional images

    Renata Khasanova and Pascal Frossard. Graph-based classifi- cation of omnidirectional images. In ICCV, 2017. 3

    2017

  11. [19]

    Piccolo: Point cloud-centric omnidirectional localiza- tion

    Junho Kim, Changwoon Choi, Hojun Jang, and Young Min Kim. Piccolo: Point cloud-centric omnidirectional localiza- tion. In CVPR, 2021. 5, 7, 12

    2021

  12. [20]

    Odgs: 3d scene reconstruction from omnidirectional images with 3d gaussian splattings

    Suyoung Lee, Jaeyoung Chung, Jaeyoo Huh, and Kyoung Mu Lee. Odgs: 3d scene reconstruction from omnidirectional images with 3d gaussian splattings. In NeurIPS, 2024. 2, 5, 6, 7, 11

    2024

  13. [21]

    Dngaussian: Optimizing sparse-view 3d gaussian radiance fields with global-local depth normalization

    Jiahe Li, Jiawei Zhang, Xiao Bai, Jin Zheng, Xin Ning, Jun Zhou, and Lin Gu. Dngaussian: Optimizing sparse-view 3d gaussian radiance fields with global-local depth normalization. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 20775–...

    2024

  14. [22]

    Vastgaussian: Vast 3d gaussians for large scene reconstruction

    Jiaqi Lin, Zhihao Li, Xiao Tang, Jianzhuang Liu, Shiyong Liu, Jiayue Liu, Yangdi Lu, Xiaofei Wu, Songcen Xu, Youliang Yan, and Wenming Yang. Vastgaussian: Vast 3d gaussians for large scene reconstruction. In CVPR, 2024. 2

    2024

  15. [23]

    Scaffold-gs: Structured 3d gaussians for view-adaptive rendering

    Tao Lu, Mulin Yu, Linning Xu, Yuanbo Xiangli, Limin Wang, Dahua Lin, and Bo Dai. Scaffold-gs: Structured 3d gaussians for view-adaptive rendering. In CVPR, 2024. 2

    2024

  16. [24]

    Srinivasan, Matthew Tancik, Jonathan T

    Ben Mildenhall, Pratul P. Srinivasan, Matthew Tancik, Jonathan T. Barron, Ravi Ramamoorthi, and Ren Ng. Nerf: Representing scenes as neural radiance fields for view synthe- sis. In ECCV, 2020. 2

    2020

  17. [25]

    Salnet360: Saliency maps for omni-directional im- ages with cnn

    Rafael Monroy, Sebastian Lutz, Tejo Chalasani, and Aljosa Smolic. Salnet360: Saliency maps for omni-directional im- ages with cnn. Signal Processing: Image Communication ,

  18. [26]

    OpenMVG: Open multiple view geometry

    Pierre Moulon, Pascal Monasse, Romuald Perrot, and Renaud Marlet. OpenMVG: Open multiple view geometry. In RRPR,

  19. [27]

    Coherentgs: Sparse novel view synthesis with coherent 3d gaussians

    Avinash Paliwal, Wei Ye, Jinhui Xiong, Dmytro Kotovenko, Rakesh Ranjan, Vikas Chandra, and Nima Khademi Kalantari. Coherentgs: Sparse novel view synthesis with coherent 3d gaussians. In ECCV, 2024. 2

    2024

  20. [28]

    Very deep convolu- tional networks for large-scale image recognition

    Karen Simonyan and Andrew Zisserman. Very deep convolu- tional networks for large-scale image recognition. In ICLR,

  21. [29]

    Learning spherical con- volution for fast features from 360 ° imagery

    Yu-Chuan Su and Kristen Grauman. Learning spherical con- volution for fast features from 360 ° imagery. In NeurIPS,

  22. [30]

    Splatter image: Ultra-fast single-view 3d reconstruc- tion

    Stanislaw Szymanowicz, Chrisitian Rupprecht, and Andrea Vedaldi. Splatter image: Ultra-fast single-view 3d reconstruc- tion. In CVPR, 2024. 2

    2024

  23. [31]

    Image quality assessment: From error visibility to structural similarity

    Zhou Wang, Alan C Bovik, Hamid R Sheikh, Eero P Simon- celli, et al. Image quality assessment: From error visibility to structural similarity. IEEE TIP, 2004. 5 9

    2004

  24. [32]

    latentsplat: Autoencoding variational gaus- sians for fast generalizable 3d reconstruction

    Christopher Wewer, Kevin Raj, Eddy Ilg, Bernt Schiele, and Jan Eric Lenssen. latentsplat: Autoencoding variational gaus- sians for fast generalizable 3d reconstruction. In ECCV, 2024. 1, 2, 5, 6, 8

    2024

  25. [33]

    Physgaussian: Physics- integrated 3d gaussians for generative dynamics

    Tianyi Xie, Zeshun Zong, Yuxing Qiu, Xuan Li, Yutao Feng, Yin Yang, and Chenfanfu Jiang. Physgaussian: Physics- integrated 3d gaussians for generative dynamics. In CVPR,

  26. [34]

    Depth- splat: Connecting gaussian splatting and depth.arXiv preprint arXiv:2410.13862, 2024

    Haofei Xu, Songyou Peng, Fangjinhua Wang, Hermann Blum, Daniel Barath, Andreas Geiger, and Marc Pollefeys. Depth- splat: Connecting gaussian splatting and depth.arXiv preprint arXiv:2410.13862, 2024. 2, 6

    2024 arXiv

  27. [35]

    Grm: Large gaussian reconstruction model for efficient 3d recon- struction and generation

    Yinghao Xu, Zifan Shi, Wang Yifan, Hansheng Chen, Ceyuan Yang, Sida Peng, Yujun Shen, and Gordon Wetzstein. Grm: Large gaussian reconstruction model for efficient 3d recon- struction and generation. In ECCV, 2024. 2

    2024

  28. [36]

    Gaus- sian grouping: Segment and edit anything in 3d scenes

    Mingqiao Ye, Martin Danelljan, Fisher Yu, and Lei Ke. Gaus- sian grouping: Segment and edit anything in 3d scenes. In ECCV, 2024. 12

    2024

  29. [37]

    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. In CVPR, 2021. 2

    2021

  30. [38]

    Gsdf: 3dgs meets sdf for improved neural rendering and reconstruction

    Mulin Yu, Tao Lu, Linning Xu, Lihan Jiang, Yuanbo Xian- gli, and Bo Dai. Gsdf: 3dgs meets sdf for improved neural rendering and reconstruction. In NeurIPS, 2024. 2

    2024

  31. [39]

    Gs-lrm: Large recon- struction model for 3d gaussian splatting

    Kai Zhang, Sai Bi, Hao Tan, Yuanbo Xiangli, Nanxuan Zhao, Kalyan Sunkavalli, and Zexiang Xu. Gs-lrm: Large recon- struction model for 3d gaussian splatting. In ECCV, 2024. 2

    2024

  32. [40]

    Efros, Eli Shechtman, and Oliver Wang

    Richard Zhang, Phillip Isola, Alexei A. Efros, Eli Shechtman, and Oliver Wang. The unreasonable effectiveness of deep features as a perceptual metric. In CVPR, 2018. 5 10 OmniSplat: Taming Feed-Forward 3D Gaussian Splatting for Omnidirectional Images with Editable Capabilities...

    2018

  33. [2024]

    1, 2, 5, 6, 7, 8, 12

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.