Jianwei Yin

Continual Video-MLLM Adaptation over Evolving Domains

Rui Cheng, Meixing Shi, Yuxiang Cai, Jingcai Guo, Jianwei Yin, Zhi Chen

Video multimodal large language models have shown strong capability in video understanding, yet their adaptation to sequentially evolving domains remains underexplored. In real-world deployments, video data often arrives continuously from heterogeneous domains, requiring the mode…Read more