Loss function search for person re-identification

Article Computer Science, Artificial Intelligence

Deep features for person re-identification on metric learning

Wanyin Wu et al.

Summary: This study summarizes different types of features and metric learning approaches for person re-identification from a label attributes perspective. By combining advanced methods in data enhancement and feature extraction, comprehensive experiments were conducted on metric learning methods using two datasets, revealing the relationships between loss functions, deep feature space, and metric learning.

PATTERN RECOGNITION (2021)

添加到收藏夹

Article Computer Science, Artificial Intelligence

3D-GAT: 3D-Guided adversarial transform network for person re-identification in unseen domains

Hengheng Zhang et al.

Summary: Person Re-identification (ReID) has made significant progress in recent years, but faces challenges in real-world applications due to differences among cameras and datasets. The 3D-GAT network synthesizes person images in different conditions using image-to-image translation, enhancing the generalization of ReID algorithms. With augmented training data, ReID approaches can better perceive variations in appearance under different viewpoints and poses, leading to higher accuracy in unknown domains and setting new state-of-the-art results in direct transfer to other datasets.

PATTERN RECOGNITION (2021)

添加到收藏夹

Article Computer Science, Artificial Intelligence

Semi-supervised blockwisely architecture search for efficient lightweight generative adversarial network

Man Zhang et al.

Summary: The study proposes a method based on semi-supervised generative adversarial networks, using real and generated images for training to achieve semi-supervised learning. By setting layer hyperparameters to variable and flexible stacking block structures, network architecture search is achieved. Experimental results show a significant reduction in training time and improved model performance.

PATTERN RECOGNITION (2021)

添加到收藏夹

Article Computer Science, Information Systems