← Search

Yingwei Song

2 accepted papers

2026

Lightning Unified Video Editing via In-Context Sparse Attention

ICML 2026poster

Video editing has evolved toward In-Context Learning (ICL) paradigms, yet the resulting quadratic attention costs create a critical computational bottleneck. In this work, we propose **I**n-context **S**parse **A**ttention (**ISA**), the first experimentally lossless sparse framework tailored for IC…

Cited by 0SourceScholar
2025

4D LangSplat: 4D Language Gaussian Splatting via Multimodal Large Language Models

CVPR 2025poster

Learning 4D language fields to enable time-sensitive, open-ended language queries in dynamic scenes is essential for many real-world applications. While LangSplat successfully grounds CLIP features into 3D Gaussian representations, achieving precision and efficiency in 3D static scenes, it lacks the…