
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.


coremsg

domain-generalizable-person-search-using-automatically-labeled-unreal-dataset


Domain Generalizable Person Search Using Automatically Labeled Unreal Dataset


title_rewrite


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.


leveraging-large-language-models-for-accurate-skeleton-based-action-recognition


Leveraging Large Language Models for Accurate Skeleton-based Action Recognition
