Enhancing Person Re-Identification through Attention-Driven Global Features and Angular Loss Optimization

To address challenges related to the inadequate representation and inaccurate discrimination of pedestrian attributes, we propose a novel method for person re-identification, which leverages global feature learning and classification optimization. Specifically, this approach integrates a Normalizati...

وصف كامل

التفاصيل البيبلوغرافية

المؤلفون الرئيسيون:	Yihan Bi, Rong Wang, Qianli Zhou, Ronghui Lin, Mingjie Wang
التنسيق:	Artigo
اللغة:	Inglês
منشور في:	MDPI AG 2024-05-01
سلاسل:	Entropy
الموضوعات:	person re-identification global feature learning attention mechanism classification loss optimization
الوصول للمادة أونلاين:	https://www.mdpi.com/1099-4300/26/6/436
الوسوم:	إضافة وسم لا توجد وسوم, كن أول من يضع وسما على هذه التسجيلة!