登录
洞察 - Artificial Intelligence Computer Vision
Domain Generalizable Person Search Using Automatically Labeled Unreal Dataset
A novel framework for domain generalizable person search that uses an automatically labeled unreal dataset for training, alleviating the need for time-consuming and labor-intensive data labeling as well as privacy issues in real datasets.
Leveraging Large Language Models for Accurate Skeleton-based Action Recognition
Large language models can be effectively leveraged as powerful action recognizers by projecting skeleton sequences into "action sentences" that are compatible with the models' pre-trained knowledge.
