21 papers
Efficient Video Dataset Distillation via Cluster-Guided Prototype Blending
Chongle Ren, Guang Li, Wenbo Huang +3
Video dataset distillation aims to compress a large video dataset into a compact surrogate set that preserves its training utility. Most existing approaches synthesize condensed vi…
FD: A Dedicated Framework for Fine-Grained Dataset Distillation
Hongxu Ma, Guang Li, Shijie Wang +5
Dataset distillation (DD) compresses a large training set into a small synthetic set, reducing storage and training cost, and has shown strong results on general benchmarks. Decoup…
L2R: Low-Rank and Lipschitz-Controlled Routing for Mixture-of-Experts
Minghao Yang, Ren Togo, Guang Li +2
Mixture-of-Experts (MoE) models scale neural networks by conditionally activating a small subset of experts, where the router plays a central role in determining expert specializat…
Predictive but Not Plannable: RC-aux for Latent World Models
Wenyuan Li, Guang Li, Keisuke Maeda +2
A latent world model may achieve accurate short-horizon prediction while still inducing a latent space that is poorly aligned with planning. A key issue is spatiotemporal mismatch:…
Closed-Form Linear-Probe Dataset Distillation for Pre-trained Vision Models
Bincheng Peng, Guang Li, Ping Liu +2
Dataset distillation compresses a large training set into a small synthetic set that preserves downstream training utility. While most existing methods target training networks fro…
GIFT: Global Irreplaceability Frame Targeting for Efficient Video Understanding
Junpeng Ma, Sashuai Zhou, Guanghao Li +9
Video Large Language Models (VLMs) have achieved remarkable success in video understanding, but the significant computational cost from processing dense frames severely limits thei…