Saved in:
Bibliographic Details
Main Authors: Gu, Xulin, Zhong, Xinhao, Wei, Zhixing, Zhou, Yimin, Sun, Shuoyang, Chen, Bin, Wang, Hongpeng, Luo, Yuan
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2505.20694
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908380680945664
author Gu, Xulin
Zhong, Xinhao
Wei, Zhixing
Zhou, Yimin
Sun, Shuoyang
Chen, Bin
Wang, Hongpeng
Luo, Yuan
author_facet Gu, Xulin
Zhong, Xinhao
Wei, Zhixing
Zhou, Yimin
Sun, Shuoyang
Chen, Bin
Wang, Hongpeng
Luo, Yuan
contents Dataset distillation (DD) has emerged as a powerful paradigm for dataset compression, enabling the synthesis of compact surrogate datasets that approximate the training utility of large-scale ones. While significant progress has been achieved in distilling image datasets, extending DD to the video domain remains challenging due to the high dimensionality and temporal complexity inherent in video data. Existing video distillation (VD) methods often suffer from excessive computational costs and struggle to preserve temporal dynamics, as naïve extensions of image-based approaches typically lead to degraded performance. In this paper, we propose a novel uni-level video dataset distillation framework that directly optimizes synthetic videos with respect to a pre-trained model. To address temporal redundancy and enhance motion preservation, we introduce a temporal saliency-guided filtering mechanism that leverages inter-frame differences to guide the distillation process, encouraging the retention of informative temporal cues while suppressing frame-level redundancy. Extensive experiments on standard video benchmarks demonstrate that our method achieves state-of-the-art performance, bridging the gap between real and distilled video data and offering a scalable solution for video dataset compression.
format Preprint
id arxiv_https___arxiv_org_abs_2505_20694
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Temporal Saliency-Guided Distillation: A Scalable Framework for Distilling Video Datasets
Gu, Xulin
Zhong, Xinhao
Wei, Zhixing
Zhou, Yimin
Sun, Shuoyang
Chen, Bin
Wang, Hongpeng
Luo, Yuan
Computer Vision and Pattern Recognition
Machine Learning
Dataset distillation (DD) has emerged as a powerful paradigm for dataset compression, enabling the synthesis of compact surrogate datasets that approximate the training utility of large-scale ones. While significant progress has been achieved in distilling image datasets, extending DD to the video domain remains challenging due to the high dimensionality and temporal complexity inherent in video data. Existing video distillation (VD) methods often suffer from excessive computational costs and struggle to preserve temporal dynamics, as naïve extensions of image-based approaches typically lead to degraded performance. In this paper, we propose a novel uni-level video dataset distillation framework that directly optimizes synthetic videos with respect to a pre-trained model. To address temporal redundancy and enhance motion preservation, we introduce a temporal saliency-guided filtering mechanism that leverages inter-frame differences to guide the distillation process, encouraging the retention of informative temporal cues while suppressing frame-level redundancy. Extensive experiments on standard video benchmarks demonstrate that our method achieves state-of-the-art performance, bridging the gap between real and distilled video data and offering a scalable solution for video dataset compression.
title Temporal Saliency-Guided Distillation: A Scalable Framework for Distilling Video Datasets
topic Computer Vision and Pattern Recognition
Machine Learning
url https://arxiv.org/abs/2505.20694