Pith. sign in

REVIEW 11 cited by

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

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2501.12375 v3 pith:C47QNCKD submitted 2025-01-21 cs.CV cs.AI

Sili Chen , Hengkai Guo , Shengnan Zhu , Feihu Zhang , Zilong Huang , Jiashi Feng , Bingyi Kang This is my paper
classification cs.CVcs.AI
keywords depthvideoanythingvideosestimationmodeltemporalability
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

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.

Discussion (0). Sign in to comment.

Forward citations

Cited by 11 Pith papers

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

  1. UniWorld-View: Large-Baseline View Synthesis via Video Diffusion Models

    cs.CV 2026-08 conditional novelty 6.0 of 10

    UniWorld-View couples an occlusion-aware point cloud renderer with a dual-stream video diffusion model to synthesize large-baseline novel views from monocular video.

  2. X-Lens: Real-Time Metric Depth Estimation with Heterogeneous Cameras

    cs.CV 2026-07 unverdicted novelty 6.0 of 10

    X-Lens fuses arbitrary calibrated fisheye and pinhole views into real-time metric depth at 41 FPS with a 0.04B-parameter model and a new 266K-frame synthetic dataset.

  3. Measuring 3D Spatial Geometric Consistency in Dynamic Video Generation

    cs.CV 2026-03 conditional novelty 6.0 of 10

    SGC quantifies 3D geometric consistency of generated videos by measuring divergence among local camera poses estimated only on static background sub-regions.

  4. Feedback Matters: Augmenting Autonomous Dissection with Visual and Topological Feedback

    cs.RO 2025-10 conditional novelty 6.0 of 10

    A stretch-based tissue connectivity estimator plus an exposure-maximizing controller and recovery planner raised autonomous dissection success on a da Vinci robot to 80%.

  5. Generative Visual Foresight Meets Task-Agnostic Pose Estimation in Robotic Table-Top Manipulation

    cs.RO 2025-08 conditional novelty 6.0 of 10

    GVF-TAPE predicts future RGB-D frames from an image and text, then extracts end-effector poses to control a robot, achieving strong success rates without action-labeled data.

  6. SpatialTrackerV2: 3D Point Tracking Made Easy

    cs.CV 2025-07 conditional novelty 6.0 of 10

    A single feed-forward model jointly estimates video depth, camera poses, and 3D point trajectories from monocular video, setting a new state of the art on TAPVid-3D.

  7. RCG: Safety-Critical Scenario Generation for Robust Autonomous Driving via Real-World Crash Grounding

    cs.RO 2025-07 conditional novelty 6.0 of 10

    RCG replaces handcrafted adversarial scenario scoring with a crash-grounded embedding and k-NN selection, yielding a 9.2% average relative improvement in ego success.

  8. RaCalNet: Radar Calibration Network for Sparse-Supervised Metric Depth Estimation

    cs.CV 2025-06 reject novelty 6.0 of 10

    A radar-camera depth estimation framework that recalibrates sparse radar points and aligns a frozen monocular depth model using sparse LiDAR labels, claiming state-of-the-art accuracy with roughly 1% supervision density.

  9. RoboScape: Physics-informed Embodied World Model

    cs.CV 2025-06 conditional novelty 5.0 of 10

    RoboScape jointly learns RGB video, depth, and keypoint-token consistency in one autoregressive world model, improving video quality, geometry, action control, synthetic-data policy training, and policy evaluation for...

  10. Reconstructing 4D Spatial Intelligence: A Survey

    cs.CV 2025-07 accept novelty 4.0 of 10

    A review that classifies 4D scene reconstruction methods into five progressive levels: low-level cues, scene components, dynamic scenes, interactions, and physics.

  11. MM-Gesture: Towards Precise Micro-Gesture Recognition through Multimodal Fusion

    cs.CV 2025-07 conditional novelty 4.0 of 10

    Combining joint, limb, RGB, Taylor-video, optical-flow, and depth streams with two video backbones and a validation-tuned weighted ensemble reaches 73.213% top-1 accuracy on iMiGUE, the best MiGA challenge result to date.

Pith tools