核心概念
A unified CLEAR model utilizing cross-transformers and pre-trained language models achieves state-of-the-art performance in person attribute recognition and retrieval tasks.
摘要
The study introduces CLEAR, a unified network for person attribute recognition and retrieval tasks. It leverages cross-transformers and pre-trained language models to address both tasks efficiently. The study demonstrates the effectiveness of the CLEAR model on five benchmarks, achieving competitive results and outperforming other models. The model's architecture, training strategy, and evaluation results are detailed, showcasing its superior performance.
统计
CLEAR 모델은 상태-of-the-art 성능을 달성합니다.
CLEAR는 cross-transformers와 사전 훈련된 언어 모델을 활용합니다.
CLEAR는 5개의 벤치마크에서 효과적인 결과를 보여줍니다.
引用
"In this study, we demonstrate that if there is a sufficiently robust network to solve person attribute recognition, it can be adapted to facilitate better performance for the retrieval task."
"The unified CLEAR model is evaluated on five benchmarks: PETA, PA100K, Market-1501, RAPv2, and UPAR-2024."