Focusing on cutting‑edge artificial‑intelligence topics, this seminar centered on three main themes: “Empowering Large‑scale Models, Multi‑modal Fusion, and Technology Implementation”. It systematically presented the latest advances and cross‑field applications in visual intelligence. On the theoretical side, it discussed core technical breakthroughs including the construction and transfer of vision large‑models (e.g. efficient network architectures, heterogeneous federated learning), multi‑modal perception and generation (video generation, 3D Gaussian Splatting, neuromorphic cameras), and embodied‑intelligent systems (world models, autonomous‑driving decision‑making). Special emphasis was placed on theoretical innovations combining large‑models with causal reasoning and value alignment. Application scenarios covered smart healthcare, intelligent interpretation of remote‑sensing images, ocean exploration, artistic creation and other vertical domains, highlighting the trend of industrial empowerment by AI technology. The content included both mature scenarios such as autonomous driving and medical imaging, as well as forward‑looking emerging fields such as neuromorphic cameras, showing the whole‑chain innovation landscape of AI from fundamental theoretical breakthroughs to industrial‑ecosystem construction.

Participant: Lidong Zheng

Participant: Yichi Zhang

Participant: Mingjian Zhang