E-mail: [email protected]
-
Few-Shot Learning: This encompasses Zero-Shot Learning, cross-domain transfer of pre-trained models, fine-tuning of visual-linguistic models under few-shot scenarios, multimodal learning, and context-aware learning, among others. (Primarily aimed at graduate students in Electronics and Information disciplines.)
-
Matrix Factorization: This involves the improvement and application of models such as Non-negative Matrix Factorization, Low-Rank Sparse Matrix Factorization, online incremental decomposition, multi-view factorization, semantic component decomposition, and more. The focus is on optimizing these models and applying them in various contexts. (Primarily directed at graduate students in Computational Mathematics.)
-
Application of Artificial Intelligence Models: This includes lightweight applications of visual perception models, such as image classification, semantic segmentation, and object detection, with an emphasis on deploying these lightweight models in fields like Agricultural Engineering and Smart Healthcare for undergraduate innovation projects and graduation designs.