default author photo

Mingyu Sun

Nanchang Hangkong University

Nanchang

China

SCHOLARLY PAPERS

1

DOWNLOADS

54

TOTAL CITATIONS

0

Scholarly Papers (1)

1.

DualFocus: A Framework for Efficient Decoupling of Visual Representation and Focus Fusion in Video LLM

Number of pages: 16 Posted: 04 Dec 2025
Nanchang Hangkong University, Nanchang Hangkong University, Chongqing University, Northwestern Polytechnic University (NPU), affiliation not provided to SSRN, National University of Defense Technology and Nanchang Hangkong University
Downloads 54 (1,031,043)

Abstract:

Loading...

Multimodal Large Language ModelVisual Question AnsweringVideo Question AnsweringVision Language ModelVideo Large Language Model