Librarian Bot. I found the following papers similar to this paper.

The following papers were recommended by the Semantic Scholar API

Can Video Diffusion Model Reconstruct 4D Geometry? (2025)
TrajectoryCrafter: Redirecting Camera Trajectory for Monocular Videos via Diffusion Models (2025)
DiST-4D: Disentangled Spatiotemporal Diffusion with Metric Depth for 4D Driving Scene Generation (2025)
Aether: Geometric-Aware Unified World Modeling (2025)
Seeing the Future, Perceiving the Future: A Unified Driving World Model for Future Generation and Perception (2025)
MuDG: Taming Multi-modal Diffusion with Gaussian Splatting for Urban Scene Reconstruction (2025)
Jasmine: Harnessing Diffusion Prior for Self-supervised Depth Estimation (2025)

Please give a thumbs up to this comment if you found it helpful!

If you want recommendations for any Paper on Hugging Face checkout this Space

You can directly ask Librarian Bot for paper recommendations by tagging it in a comment: \n\n@librarian-bot\n\t recommend

\n","updatedAt":"2025-04-03T01:33:49.213Z","author":{"_id":"63d3e0e8ff1384ce6c5dd17d","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/1674830754237-63d3e0e8ff1384ce6c5dd17d.jpeg","fullname":"Librarian Bot (Bot)","name":"librarian-bot","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":318,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.6748038530349731},"editors":["librarian-bot"],"editorAvatarUrls":["https://cdn-avatars.huggingface.co/v1/production/uploads/1674830754237-63d3e0e8ff1384ce6c5dd17d.jpeg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2504.01016","authors":[{"_id":"67ec958ebb1d6dd924f94a31","user":{"_id":"65f8e4778dc7bb5b4db97f92","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/no-auth/gBmQdovmANmjV72k3gaW8.png","isPro":false,"fullname":"Tian-Xing Xu","user":"slothfulxtx","type":"user"},"name":"Tian-Xing Xu","status":"admin_assigned","statusLastChangedAt":"2025-04-02T09:51:06.226Z","hidden":false},{"_id":"67ec958ebb1d6dd924f94a32","user":{"_id":"64c0953a8137192a1e2474dc","avatarUrl":"/avatars/546405a7eaf2f60ad108ceaa0dda7d08.svg","isPro":false,"fullname":"xiangjun gao","user":"xiangjun0211","type":"user"},"name":"Xiangjun Gao","status":"admin_assigned","statusLastChangedAt":"2025-04-02T09:51:14.978Z","hidden":false},{"_id":"67ec958ebb1d6dd924f94a33","user":{"_id":"657a7458afbb0117ba15c59f","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/657a7458afbb0117ba15c59f/8_iwTS1UG_mKnfylFbLsY.jpeg","isPro":false,"fullname":"Wenbo Hu","user":"wbhu-tc","type":"user"},"name":"Wenbo Hu","status":"claimed_verified","statusLastChangedAt":"2025-04-02T08:22:52.313Z","hidden":false},{"_id":"67ec958ebb1d6dd924f94a34","user":{"_id":"66d8284b6bddfb32e77ddafb","avatarUrl":"/avatars/edd371ecef6e7d58b945a30bbc6095ee.svg","isPro":false,"fullname":"Xiaoyu Li","user":"Xiaoyu521","type":"user"},"name":"Xiaoyu Li","status":"claimed_verified","statusLastChangedAt":"2025-09-03T09:06:53.629Z","hidden":false},{"_id":"67ec958ebb1d6dd924f94a35","name":"Song-Hai Zhang","hidden":false},{"_id":"67ec958ebb1d6dd924f94a36","user":{"_id":"63ca3ddc04c979828310bfcb","avatarUrl":"/avatars/615e0d8622950b4408b40d550f02a894.svg","isPro":false,"fullname":"Ying Shan","user":"yshan2u","type":"user"},"name":"Ying Shan","status":"admin_assigned","statusLastChangedAt":"2025-04-02T09:51:45.869Z","hidden":false}],"mediaUrls":["https://cdn-uploads.huggingface.co/production/uploads/657a7458afbb0117ba15c59f/6He3mQcB_AO1G_Sq8xYc0.mp4"],"publishedAt":"2025-04-01T17:58:03.000Z","submittedOnDailyAt":"2025-04-02T00:15:17.585Z","title":"GeometryCrafter: Consistent Geometry Estimation for Open-world Videos\n with Diffusion Priors","submittedOnDailyBy":{"_id":"657a7458afbb0117ba15c59f","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/657a7458afbb0117ba15c59f/8_iwTS1UG_mKnfylFbLsY.jpeg","isPro":false,"fullname":"Wenbo Hu","user":"wbhu-tc","type":"user"},"summary":"Despite remarkable advancements in video depth estimation, existing methods\nexhibit inherent limitations in achieving geometric fidelity through the\naffine-invariant predictions, limiting their applicability in reconstruction\nand other metrically grounded downstream tasks. We propose GeometryCrafter, a\nnovel framework that recovers high-fidelity point map sequences with temporal\ncoherence from open-world videos, enabling accurate 3D/4D reconstruction,\ncamera parameter estimation, and other depth-based applications. At the core of\nour approach lies a point map Variational Autoencoder (VAE) that learns a\nlatent space agnostic to video latent distributions for effective point map\nencoding and decoding. Leveraging the VAE, we train a video diffusion model to\nmodel the distribution of point map sequences conditioned on the input videos.\nExtensive evaluations on diverse datasets demonstrate that GeometryCrafter\nachieves state-of-the-art 3D accuracy, temporal consistency, and generalization\ncapability.","upvotes":29,"discussionId":"67ec9593bb1d6dd924f94b3e","projectPage":"https://geometrycrafter.github.io/","githubRepo":"https://github.com/TencentARC/GeometryCrafter","githubRepoAddedBy":"user","ai_summary":"GeometryCrafter uses a point map VAE and video diffusion model to estimate high-fidelity, temporally coherent depth maps from open-world videos, enhancing 3D reconstruction and camera parameter estimation.","ai_keywords":["point map VAE","video diffusion model","high-fidelity point maps","temporal coherence","3D/4D reconstruction","camera parameter estimation","latent space","distribution modeling"],"githubStars":431},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"657a7458afbb0117ba15c59f","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/657a7458afbb0117ba15c59f/8_iwTS1UG_mKnfylFbLsY.jpeg","isPro":false,"fullname":"Wenbo Hu","user":"wbhu-tc","type":"user"},{"_id":"67e9fc3797cd6860c81d5838","avatarUrl":"/avatars/6c37731156bf52c123bd390823890d28.svg","isPro":false,"fullname":"Jangho Park","user":"jhpark96","type":"user"},{"_id":"620783f24e28382272337ba4","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/620783f24e28382272337ba4/zkUveQPNiDfYjgGhuFErj.jpeg","isPro":false,"fullname":"GuoLiangTang","user":"Tommy930","type":"user"},{"_id":"65f8e4778dc7bb5b4db97f92","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/no-auth/gBmQdovmANmjV72k3gaW8.png","isPro":false,"fullname":"Tian-Xing Xu","user":"slothfulxtx","type":"user"},{"_id":"6342796a0875f2c99cfd313b","avatarUrl":"/avatars/98575092404c4197b20c929a6499a015.svg","isPro":false,"fullname":"Yuseung \"Phillip\" Lee","user":"phillipinseoul","type":"user"},{"_id":"635964636a61954080850e1d","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/635964636a61954080850e1d/0bfExuDTrHTtm8c-40cDM.png","isPro":false,"fullname":"William Lamkin","user":"phanes","type":"user"},{"_id":"63c5d43ae2804cb2407e4d43","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/1673909278097-noauth.png","isPro":false,"fullname":"xziayro","user":"xziayro","type":"user"},{"_id":"6506b77a773ceaa8d52ecea1","avatarUrl":"/avatars/0e769a0795063e1491c44760a4a83097.svg","isPro":false,"fullname":"CJH","user":"Howe666","type":"user"},{"_id":"638f308fc4444c6ca870b60a","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/638f308fc4444c6ca870b60a/Q11NK-8-JbiilJ-vk2LAR.png","isPro":true,"fullname":"Linoy Tsaban","user":"linoyts","type":"user"},{"_id":"66fa882fa9312392f2d5a6e0","avatarUrl":"/avatars/b4fe2141925bc187636f2effa45d0cba.svg","isPro":false,"fullname":"BeeFX","user":"BeeFX","type":"user"},{"_id":"648eb1eb59c4e5c87dc116e0","avatarUrl":"/avatars/c636cea39c2c0937f01398c94ead5dad.svg","isPro":false,"fullname":"fdsqefsgergd","user":"T-representer","type":"user"},{"_id":"634dffc49b777beec3bc6448","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/1670144568552-634dffc49b777beec3bc6448.jpeg","isPro":false,"fullname":"Zhipeng Yang","user":"svjack","type":"user"}],"acceptLanguages":["*"],"dailyPaperRank":0}">

Papers

arxiv:2504.01016

GeometryCrafter: Consistent Geometry Estimation for Open-world Videos with Diffusion Priors

Published on Apr 1, 2025

· Submitted by

Wenbo Hu on Apr 2, 2025

Upvote

Authors:

Tian-Xing Xu ,

Xiangjun Gao ,

Wenbo Hu ,

Xiaoyu Li ,

Ying Shan

Abstract

GeometryCrafter uses a point map VAE and video diffusion model to estimate high-fidelity, temporally coherent depth maps from open-world videos, enhancing 3D reconstruction and camera parameter estimation.

AI-generated summary

Despite remarkable advancements in video depth estimation, existing methods exhibit inherent limitations in achieving geometric fidelity through the affine-invariant predictions, limiting their applicability in reconstruction and other metrically grounded downstream tasks. We propose GeometryCrafter, a novel framework that recovers high-fidelity point map sequences with temporal coherence from open-world videos, enabling accurate 3D/4D reconstruction, camera parameter estimation, and other depth-based applications. At the core of our approach lies a point map Variational Autoencoder (VAE) that learns a latent space agnostic to video latent distributions for effective point map encoding and decoding. Leveraging the VAE, we train a video diffusion model to model the distribution of point map sequences conditioned on the input videos. Extensive evaluations on diverse datasets demonstrate that GeometryCrafter achieves state-of-the-art 3D accuracy, temporal consistency, and generalization capability.