TAPVid-MV: A Benchmark for Tracking Any Point in 3D Across Multiple Views
cs.CV
Submitted: 2026-09-01
Updated: 2026-10-09
Project page: https://tapvidmv.github.io/dataviewer
Terminology
Sources
- Virtual KITTI 2
- SAM 3: Segment Anything with Concepts
- Look Around and Pay Attention: Multi-camera Point Tracking Reimagined with Transformers
- Syn4D: A Multiview Synthetic 4D Dataset
- TAPNext++: What's Next for Tracking Any Point (TAP)?
- CoTracker3: Simpler and Better Point Tracking by Pseudo-Labelling Real Videos
- MV-TAP: Tracking Any Point in Multi-View Videos
- Kimodo: Scaling Controllable Human Motion Generation
- THFM: A Unified Video Foundation Model for 4D Human Perception and Beyond
- SpatialTrackerV2: 3D Point Tracking Made Easy
- SynthVerse: A Large-Scale Diverse Synthetic Dataset for Point Tracking
Related papers
- Loss Knows Best: Detecting Annotation Errors in Videos via Loss Trajectories
- AnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memories
- Benchmarking the Robustness of Foundation Models for Mammography under Domain Shift
- MambaX-Net: Dual-Input Mamba-Enhanced Cross-Attention Network for Longitudinal MRI Segmentation
- TeleOCR: Navigating Document Parsing Across Digital and Camera-Captured Documents
- A Survey on Efficient Vision-Language-Action Models