mlsanityData to 7 Oct 2026

DINO-Tracker: Taming DINO for Self-Supervised Point Tracking in a Single Video

arXiv 2403.14548 ↗First posted 21 Mar 2024cs.CVcited by 64

Not on the map. Why?

Abstract (arXiv)

We present DINO-Tracker -- a new framework for long-term dense tracking in video. The pillar of our approach is combining test-time training on a single video, with the powerful localized semantic features learned by a pre-trained DINO-ViT model. Specifically, our framework simultaneously adopts DINO's features to fit to the motion observations of the test video, while training a tracker that directly leverages the refined features. The entire framework is trained end-to-end using a combination of self-supervised losses, and regularization that allows us to retain and benefit from DINO's semantic prior. Extensive evaluation demonstrates that our method achieves state-of-the-art results on known benchmarks. DINO-tracker significantly outperforms self-supervised methods and is competitive with state-of-the-art supervised trackers, while outperforming them in challenging cases of tracking under long-term occlusions.

Key facts. Cited by 64 papers in the map (data through 7 Oct 2026). Builds on 45 of its 64 references that are in the map. First posted on arXiv 21 Mar 2024.

Builds on

45 of 64 references.

References · 45

  1. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  2. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 29,915
  3. 2021LoRA: Low-Rank Adaptation of Large Language Modelscited by 14,084
  4. 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 11,979
  5. 2023DINOv2: Learning Robust Visual Features without Supervisioncited by 7,764
  6. 2021Emerging Properties in Self-Supervised Vision Transformerscited by 6,893
  7. 2023Adding Conditional Control to Text-to-Image Diffusion Modelscited by 5,466
  8. 2017Quo Vadis, Action Recognition? A New Model and the Kinetics Datasetcited by 3,259
  9. 2020RAFT: Recurrent All-Pairs Field Transforms for Optical Flowcited by 2,385
  10. 2015FlowNet: Learning Optical Flow with Convolutional Networkscited by 1,054
  11. 2017PWC-Net: CNNs for Optical Flow Using Pyramid, Warping, and Cost Volumecited by 966
  12. 2017The 2017 DAVIS Challenge on Video Object Segmentationcited by 942
33 more
  1. 2016FlowNet 2.0: Evolution of Optical Flow Estimation with Deep Networkscited by 780
  2. 2023CoTracker: It is Better to Track Togethercited by 516
  3. 2019Making Convolutional Networks Shift-Invariant Againcited by 390
  4. 2021Deep ViT Features as Dense Visual Descriptorscited by 327
  5. 2021GMFlow: Learning Optical Flow via Global Matchingcited by 324
  6. 2023A Tale of Two Features: Stable Diffusion Complements DINO for Zero-Shot Semantic Correspondencecited by 297
  7. 2023TAPIR: Tracking Any Point with per-frame Initialization and temporal Refinementcited by 281
  8. 2023PointOdyssey: A Large-Scale Synthetic Dataset for Long-Term Point Trackingcited by 274
  9. 2022FlowFormer: A Transformer Architecture for Optical Flowcited by 273
  10. 2019Learning Correspondence from the Cycle-Consistency of Timecited by 269
  11. 2022TAP-Vid: A Benchmark for Tracking Any Point in a Videocited by 258
  12. 2022Unsupervised Semantic Segmentation by Distilling Feature Correspondencescited by 241
  13. 2020Space-Time Correspondence as a Contrastive Random Walkcited by 239
  14. 2022Particle Video Revisited: Tracking Through Occlusions Using Point Trajectoriescited by 215
  15. 2022Splicing ViT Features for Semantic Appearance Transfercited by 209
  16. 2023Tracking Everything Everywhere All at Oncecited by 190
  17. 2018Tracking Emerges by Colorizing Videoscited by 166
  18. 2022Deep Spectral Methods: A Surprisingly Strong Baseline for Unsupervised Semantic Segmentation and Localizationcited by 150
  19. 2020Learning Feature Descriptors using Camera Pose Supervisioncited by 104
  20. 2016Learning Dense Correspondence via 3D-guided Cycle Consistencycited by 91
  21. 2022ParticleSfM: Exploiting Dense Point Trajectories for Localizing Moving Cameras in the Wildcited by 90
  22. 2023Telling Left from Right: Identifying Geometry-Aware Semantic Correspondencecited by 85
  23. 2018Creatures great and SMAL: Recovering the shape and motion of animals from videocited by 84
  24. 2021Rethinking Self-supervised Correspondence Learning: A Video Frame-level Similarity Perspectivecited by 81
  25. 2019Joint-task Self-supervised Learning for Temporal Correspondencecited by 80
  26. 2017Accurate Optical Flow via Direct Cost Volume Processingcited by 66
  27. 2023MFT: Long-Term Tracking of Every Pixelcited by 61
  28. 2022Learning Pixel Trajectories with Multiscale Contrastive Random Walkscited by 36
  29. 2023Time Does Tell: Self-Supervised Time-Tuning of Dense Image Representationscited by 35
  30. 2023Improving Semantic Correspondence with Viewpoint-Guided Spherical Mapscited by 33
  31. 2023Neural Congealing: Aligning Images to a Joint Semantic Atlascited by 30
  32. 2022DeepCut: Unsupervised Segmentation using Graph Neural Networks Clusteringcited by 25
  33. 2023Disentangling Structure and Appearance in ViT Feature Spacecited by 5

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 15; 2025: 31; 2026: 18.

64 papers.

Sort:

By area:

  1. 2024MoSca: Dynamic Gaussian Fusion from Casual Videos via 4D Motion Scaffoldscited by 177
  2. 2024DINOv2 Meets Text: A Unified Framework for Image- and Pixel-Level Vision-Language Alignmentcited by 79
  3. 2024DELTA: Dense Efficient Long-range 3D Tracking for any videocited by 62
  4. 2024Track4Gen: Teaching Video Diffusion Models to Track Points Improves Video Generationcited by 42
  5. 2025TAPNext: Tracking Any Point (TAP) as Next Token Predictioncited by 35
  6. 2024Beyond Bare Queries: Open-Vocabulary Object Grounding with 3D Scene Graphcited by 32
  7. 2024Feat2GS: Probing Visual Foundation Models with Gaussian Splattingcited by 26
  8. 2025LoftUp: Learning a Coordinate-Based Feature Upsampler for Vision Foundation Modelscited by 23
  9. 2025JAFAR: Jack up Any Feature at Any Resolutioncited by 21
  10. 2024Retrieval-augmented Few-shot Medical Image Segmentation with Foundation Modelscited by 20
  11. 2024Zero-Shot Object-Centric Representation Learningcited by 19
  12. 2024When Does Perceptual Alignment Benefit Vision Representations?cited by 18
38 more
  1. 2025Cluster and Predict Latent Patches for Improved Masked Image Modelingcited by 17
  2. 2025Exploring Temporally-Aware Features for Point Trackingcited by 15
  3. 2024Multiview Equivariance Improves 3D Correspondence Understanding with Minimal Feature Finetuningcited by 13
  4. 2025Seurat: From Moving Points to Depthcited by 13
  5. 2025Time-to-Move: Training-Free Motion Controlled Video Generation via Dual-Clock Denoisingcited by 13
  6. 2024EgoPoints: Advancing Point Tracking for Egocentric Videoscited by 10
  7. 2025Geometry-aware RL for Manipulation of Varying Shapes and Deformable Objectscited by 10
  8. 2025Temporal Unlearnable Examples: Preventing Personal Video Data from Unauthorized Exploitation by Object Trackingcited by 9
  9. 2025DELTAv2: Accelerating Dense 3D Trackingcited by 9
  10. 2025LookOut: Real-World Humanoid Egocentric Navigationcited by 9
  11. 2025Empowering DINO Representations for Underwater Instance Segmentation via Aligner and Promptercited by 9
  12. 2024TAPTRv3: Spatial and Temporal Context Foster Robust Tracking of Any Point in Long Videocited by 8
  13. 2025Trokens: Semantic-Aware Relational Trajectory Tokens for Few-Shot Action Recognitioncited by 7
  14. 2024Tracking Any Point with Frame-Event Fusion Network at High Frame Ratecited by 6
  15. 2024TieBot: Learning to Knot a Tie from Visual Demonstration through a Real-to-Sim-to-Real Approachcited by 5
  16. 2025DINO-VO: A Feature-based Visual Odometry Leveraging a Visual Foundation Modelcited by 5
  17. 2026Video Understanding: From Geometry and Semantics to Unified Modelscited by 5
  18. 2025Self-Supervised Learning of Motion Concepts by Optimizing Counterfactualscited by 4
  19. 20253D-Aware Vision-Language Models Fine-Tuning with Geometric Distillationcited by 4
  20. 2025Look Around and Pay Attention: Multi-camera Point Tracking Reimagined with Transformerscited by 4
  21. 2025ProTracker: Probabilistic Integration for Robust and Accurate Point Trackingcited by 3
  22. 2025CleverDistiller: Simple and Spatially Consistent Cross-modal Distillationcited by 3
  23. 2025Order Matters: On Parameter-Efficient Image-to-Video Probing for Recognizing Nearly Symmetric Actionscited by 3
  24. 2025AnthroTAP: Learning Point Tracking with Real-World Motioncited by 3
  25. 2025View-Consistent Diffusion Representations for 3D-Consistent Video Generationcited by 3
  26. 2025Probing and Leveraging Video Diffusion Transformer Features for Robust Point Trackingcited by 3
  27. 2026SemanticMoments: Training-Free Motion Similarity via Third Moment Featurescited by 3
  28. 2026MARCO: Navigating the Unseen Space of Semantic Correspondencecited by 3
  29. 2025Animal Pose Labeling Using General-Purpose Point Trackerscited by 2
  30. 2025FRAME: Pre-Training Video Feature Representations via Anticipation and Memorycited by 2
  31. 2025Taming Modern Point Tracking for Speckle Tracking Echocardiography via Impartial Motioncited by 2
  32. 2025PointSt3R: Point Tracking through 3D Grounded Correspondencecited by 2
  33. 2025Multi-View Foundation Modelscited by 2
  34. 2024Kalib: Easy Hand-Eye Calibration with Reference Point Trackingcited by 1
  35. 2026HDR-NSFF: High Dynamic Range Neural Scene Flow Fieldscited by 1
  36. 2026DRoPS: Dynamic 3D Reconstruction of Pre-Scanned Objectscited by 1
  37. 2025HQ-SMem: Video Segmentation and Tracking Using Memory Efficient Object Embedding With Selective Update and Self-Supervised Distillation Feedbackcited by 0
  38. 2025Densemarks: Learning Canonical Embeddings for Human Heads Images via Point Trackscited by 0

50 most cited of 64 All 64 (JSON)