mlsanityData to 7 Oct 2026

Video Depth Anything: Consistent Depth Estimation for Super-Long Videos

arXiv 2501.12375 ↗First posted 21 Jan 2025cs.CV cs.AIcited by 202

Not on the map. Why?

Cited by 33 in Diffusion models, but placed by its own abstract, which didn’t fit.

Abstract (arXiv)

Depth Anything has achieved remarkable success in monocular depth estimation with strong generalization ability. However, it suffers from temporal inconsistency in videos, hindering its practical applications. Various methods have been proposed to alleviate this issue by leveraging video generation models or introducing priors from optical flow and camera poses. Nonetheless, these methods are only applicable to short videos (< 10 seconds) and require a trade-off between quality and computational efficiency. We propose Video Depth Anything for high-quality, consistent depth estimation in super-long videos (over several minutes) without sacrificing efficiency. We base our model on Depth Anything V2 and replace its head with an efficient spatial-temporal head. We design a straightforward yet effective temporal consistency loss by constraining the temporal depth gradient, eliminating the need for additional geometric priors. The model is trained on a joint dataset of video depth and unlabeled images, similar to Depth Anything V2. Moreover, a novel key-frame-based strategy is developed for long video inference. Experiments show that our model can be applied to arbitrarily long videos without compromising quality, consistency, or generalization ability. Comprehensive evaluations on multiple video benchmarks demonstrate that our approach sets a new state-of-the-art in zero-shot video depth estimation. We offer models of different scales to support a range of scenarios, with our smallest model capable of real-time performance at 30 FPS.

Key facts. Cited by 202 papers in the map (data through 7 Oct 2026). Builds on 36 of its 52 references that are in the map. First posted on arXiv 21 Jan 2025.

Builds on

36 of 52 references.

References · 36

  1. 2017Attention Is All You Needcited by 59,932
  2. 2023DINOv2: Learning Robust Visual Features without Supervisioncited by 7,764
  3. 2017ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenescited by 3,088
  4. 2023Stable Video Diffusion: Scaling Latent Video Diffusion Models to Large Datasetscited by 2,526
  5. 2020RAFT: Recurrent All-Pairs Field Transforms for Optical Flowcited by 2,385
  6. 2021Vision Transformers for Dense Predictioncited by 1,789
  7. 2024Depth Anything V2cited by 1,385
  8. 2024Depth Anything: Unleashing the Power of Large-Scale Unlabeled Datacited by 1,346
  9. 2014Depth Map Prediction from a Single Image using a Multi-Scale Deep Networkcited by 1,264
  10. 2018Deep Ordinal Regression Network for Monocular Depth Estimationcited by 721
  11. 2023ZoeDepth: Zero-shot Transfer by Combining Relative and Metric Depthcited by 646
  12. 2020AdaBins: Depth Estimation using Adaptive Binscited by 596
24 more
  1. 2020TartanAir: A Dataset to Push the Limits of Visual SLAMcited by 494
  2. 2024UniDepth: Universal Monocular Metric Depth Estimationcited by 375
  3. 2023Metric3D: Towards Zero-shot Metric 3D Prediction from A Single Imagecited by 357
  4. 2020Virtual KITTI 2cited by 345
  5. 2021GMFlow: Learning Optical Flow via Global Matchingcited by 324
  6. 2023ControlVideo: Training-free Controllable Text-to-Video Generationcited by 324
  7. 2024MoGe: Unlocking Accurate Monocular Geometry Estimation for Open-Domain Images with Optimal Training Supervisioncited by 290
  8. 2023PointOdyssey: A Large-Scale Synthetic Dataset for Long-Term Point Trackingcited by 274
  9. 2019DIODE: A Dense Indoor and Outdoor DEpth Datasetcited by 236
  10. 2020Consistent Video Depth Estimationcited by 220
  11. 2024DepthCrafter: Generating Consistent Long Depth Sequences for Open-world Videoscited by 202
  12. 2020Robust Consistent Video Depth Estimationcited by 188
  13. 2019ReFusion: 3D Reconstruction in Dynamic Environments for RGB-D Cameras Exploiting Residualscited by 180
  14. 2023MiDaS v3.1 -- A Model Zoo for Robust Monocular Relative Depth Estimationcited by 173
  15. 2024ControlNeXt: Powerful and Efficient Control for Image and Video Generationcited by 158
  16. 2022NeW CRFs: Neural Window Fully-connected CRFs for Monocular Depth Estimationcited by 151
  17. 2024Learning Temporally Consistent Video Depth from Video Diffusion Priorscited by 100
  18. 2022SimpleRecon: 3D Reconstruction Without 3D Convolutionscited by 83
  19. 2024Depth Any Video with Scalable Synthetic Datacited by 57
  20. 2021Consistent Depth of Moving Objects in Videocited by 56
  21. 2021Towards Real-Time Monocular Depth Estimation for Robotics: A Surveycited by 55
  22. 2022Less is More: Consistent Video Depth Estimation with Masked Frames Modelingcited by 30
  23. 2023MAMo: Leveraging Memory and Attention for Monocular Video Depth Estimationcited by 24
  24. 2024Match-Stereo-Videos: Bidirectional Alignment for Consistent Dynamic Stereo Matchingcited by 24

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 92; 2026: 110.

202 papers.

Sort:

By area:

  1. 2025World Simulation with Video Foundation Models for Physical AIcited by 183
  2. 2025UniDepthV2: Universal Monocular Metric Depth Estimation Made Simplercited by 170
  3. 2025ViPE: Video Pose Engine for 3D Geometric Perceptioncited by 162
  4. 2025SpatialTrackerV2: 3D Point Tracking Made Easycited by 84
  5. 2025Scaling Instruction-Based Video Editing with a High-Quality Synthetic Datasetcited by 76
  6. 2026Unified 4D World Action Modeling from Video Priors with Asynchronous Denoisingcited by 65
  7. 2025GigaWorld-0: World Models as Data Engine to Empower Embodied AIcited by 60
  8. 2025OpenVE-3M: A Large-Scale High-Quality Dataset for Instruction-Guided Video Editingcited by 50
  9. 2025Pixel-Perfect Depth with Semantics-Prompted Diffusion Transformerscited by 47
  10. 2025GigaBrain-0: A World Model-Powered Vision-Language-Action Modelcited by 43
  11. 2025Puppeteer: Rig and Animate Your 3D Modelscited by 41
  12. 2025RoboScape: Physics-informed Embodied World Modelcited by 38
38 more
  1. 2025Learning Video Generation for Robotic Manipulation with Collaborative Trajectory Controlcited by 33
  2. 2025FantasyWorld: Geometry-Consistent World Modeling via Unified Video and 3D Predictioncited by 33
  3. 2025Any4D: Unified Feed-Forward Metric 4D Reconstructioncited by 32
  4. 2025MoVieS: Motion-Aware 4D Dynamic View Synthesis in One Secondcited by 31
  5. 2025FlashWorld: High-quality 3D Scene Generation within Secondscited by 30
  6. 2025PAI-Bench: A Comprehensive Benchmark For Physical AIcited by 30
  7. 2026Rethinking Video Generation Model for the Embodied Worldcited by 30
  8. 2025WorldExplorer: Towards Generating Fully Navigable 3D Scenescited by 26
  9. 2025POMATO: Marrying Pointmap Matching with Temporal Motion for Dynamic 3D Reconstructioncited by 24
  10. 2025OmniVDiff: Omni Controllable Video Diffusion for Generation and Understandingcited by 22
  11. 2025LongVie: Multimodal-Guided Controllable Ultra-Long Video Generationcited by 21
  12. 2025QDepth-VLA: Quantized Depth Prediction as Auxiliary Supervision for Vision-Language-Action Modelscited by 19
  13. 2025FlowDreamer: A RGB-D World Model with Flow-based Motion Representations for Robot Manipulationcited by 17
  14. 2025Time Blindness: Why Video-Language Models Can't See What Humans Can?cited by 17
  15. 2025OmniVCus: Feedforward Subject-driven Video Customization with Multimodal Control Conditionscited by 17
  16. 2025Memorize-and-Generate: Towards Long-Term Consistency in Real-Time Video Generationcited by 15
  17. 2026RoboWM-Bench: A Benchmark for Evaluating World Models in Robotic Manipulationcited by 15
  18. 2026Vista4D: Video Reshooting with 4D Point Cloudscited by 15
  19. 2025Endo3R: Unified Online Reconstruction from Dynamic Monocular Endoscopic Videocited by 14
  20. 2025FlashDepth: Real-time Streaming Video Depth Estimation at 2K Resolutioncited by 14
  21. 2025GeoVideo: Introducing Geometric Regularization into Video Generation Modelcited by 14
  22. 2025CARI4D: Category Agnostic 4D Reconstruction of Human-Object Interactioncited by 14
  23. 2025MM-Gesture: Towards Precise Micro-Gesture Recognition through Multimodal Fusioncited by 13
  24. 2025Reconstructing 4D Spatial Intelligence: A Surveycited by 13
  25. 2025EMMA: Generalizing Real-World Robot Manipulation via Generative Visual Transfercited by 12
  26. 2025Vivid4D: Improving 4D Reconstruction from Monocular Video by Video Inpaintingcited by 11
  27. 2025DualCamCtrl: Dual-Branch Diffusion Model for Geometry-Aware Camera-Controlled Video Generationcited by 11
  28. 2025ReDirector: Creating Any-Length Video Retakes with Rotary Camera Encodingcited by 10
  29. 2026Dex4D: Task-Agnostic Point Track Policy for Sim-to-Real Dexterous Manipulationcited by 10
  30. 2025UniGeo: Taming Video Diffusion for Unified Consistent Geometry Estimationcited by 9
  31. 2025Tinker: Diffusion's Gift to 3D--Multi-View Consistent Editing From Sparse Inputs without Per-Scene Optimizationcited by 9
  32. 2025LongVie 2: Multimodal Controllable Ultra-Long Video World Modelcited by 9
  33. 2025The Fourth Monocular Depth Estimation Challengecited by 8
  34. 20254D-VGGT: A General Foundation Model with SpatioTemporal Awareness for Dynamic Scene Geometry Estimationcited by 8
  35. 2025EgoX: Egocentric Video Generation from a Single Exocentric Videocited by 8
  36. 2026Tele-Omni: a Unified Multimodal Framework for Video Generation and Editingcited by 8
  37. 2026Generative World Renderercited by 8
  38. 2025Seeing World Dynamics in a Nutshellcited by 7

50 most cited of 202 All 202 (JSON)