Repurposing Video Diffusion Transformers for Robust Point Tracking

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Son, Soowon, An, Honggyu, Kim, Chaehyun, Ko, Hyunah, Nam, Jisu, Chung, Dahyun, Jin, Siyoon, Yi, Jung, Min, Jaewon, Hur, Junhwa, Kim, Seungryong
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866909974848864256
author Son, Soowon
An, Honggyu
Kim, Chaehyun
Ko, Hyunah
Nam, Jisu
Chung, Dahyun
Jin, Siyoon
Yi, Jung
Min, Jaewon
Hur, Junhwa
Kim, Seungryong
author_facet Son, Soowon
An, Honggyu
Kim, Chaehyun
Ko, Hyunah
Nam, Jisu
Chung, Dahyun
Jin, Siyoon
Yi, Jung
Min, Jaewon
Hur, Junhwa
Kim, Seungryong
contents Point tracking aims to localize corresponding points across video frames, serving as a fundamental task for 4D reconstruction, robotics, and video editing. Existing methods commonly rely on shallow convolutional backbones such as ResNet that process frames independently, lacking temporal coherence and producing unreliable matching costs under challenging conditions. Through systematic analysis, we find that video Diffusion Transformers (DiTs), pre-trained on large-scale real-world videos with spatio-temporal attention, inherently exhibit strong point tracking capability and robustly handle dynamic motions and frequent occlusions. We propose DiTracker, which adapts video DiTs through: (1) query-key attention matching, (2) lightweight LoRA tuning, and (3) cost fusion with a ResNet backbone. Despite training with 8 times smaller batch size, DiTracker achieves state-of-the-art performance on challenging ITTO benchmark and matches or outperforms state-of-the-art models on TAP-Vid benchmarks. Our work validates video DiT features as an effective and efficient foundation for point tracking.
format Preprint
id arxiv_https___arxiv_org_abs_2512_20606
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Repurposing Video Diffusion Transformers for Robust Point Tracking
Son, Soowon
An, Honggyu
Kim, Chaehyun
Ko, Hyunah
Nam, Jisu
Chung, Dahyun
Jin, Siyoon
Yi, Jung
Min, Jaewon
Hur, Junhwa
Kim, Seungryong
Computer Vision and Pattern Recognition
Point tracking aims to localize corresponding points across video frames, serving as a fundamental task for 4D reconstruction, robotics, and video editing. Existing methods commonly rely on shallow convolutional backbones such as ResNet that process frames independently, lacking temporal coherence and producing unreliable matching costs under challenging conditions. Through systematic analysis, we find that video Diffusion Transformers (DiTs), pre-trained on large-scale real-world videos with spatio-temporal attention, inherently exhibit strong point tracking capability and robustly handle dynamic motions and frequent occlusions. We propose DiTracker, which adapts video DiTs through: (1) query-key attention matching, (2) lightweight LoRA tuning, and (3) cost fusion with a ResNet backbone. Despite training with 8 times smaller batch size, DiTracker achieves state-of-the-art performance on challenging ITTO benchmark and matches or outperforms state-of-the-art models on TAP-Vid benchmarks. Our work validates video DiT features as an effective and efficient foundation for point tracking.
title Repurposing Video Diffusion Transformers for Robust Point Tracking
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2512.20606