Siamese Network: Person Re-Identification Using Deep Learning with PyTorch

Author(s)	Ms. Gorle Dhana Lakshmi, Mr. Mudidana Jeevaratnam, Mr. Mudidana Sukumar
Country	India
Abstract	Learning is made challenging by problems with person re-identification (Re-ID), like occlusion and misalignment. To rectify these problems, it’s very essential to emphasize strong characteristics in intra-class variance. The attention-based Re-ID techniques used today ignore unique characteristics in favour of commonalities. In this research presents a novel Siamese network for person’s Re-ID on attentive learning. In contrast to prior techniques, the leveraged the Siamese network's characteristics to create the attention-loss and attention module that focus attention on recognizable and common elements. The encoder-decoder attention module is to determine the general form of the body and channel attention to choose significant channels. The term "uniformity loss" now describes the attention deficit brought on by the triplet loss. A distinct attention map that emphasizes both common and discriminative characteristics is produced because of the homogeneity loss. The recommended network performs better than the most recent techniques on the Market-1501, according to a comprehensive testing procedure.
Keywords	Re-identification, Siamese network, Market-1501 data sets.
Field	Computer > Artificial Intelligence / Simulation / Virtual Reality
Published In	Volume 8, Issue 2, March-April 2026
Published On	2026-04-14
DOI	https://doi.org/10.36948/ijfmr.2026.v08i02.73794

About IJFMR Fees & Payment Current Issue Publication Archive	Submit Research Paper Track Submission Status Publication Guidelines Publication Ethics Peer Review & Plagiarism	Join as a Reviewer Editors & Reviewers Reviewer Referral Program Get Reviewer Membership Certi.	Website/Journal Policies Usage Policy Content Policies Privacy Policy

Contact Us		+91-9687-828-838	editor@ijfmr.com

International Journal For Multidisciplinary Research