🤖 AI Summary
To address the over-parameterization, high computational cost, and deployment challenges of deep neural networks for Person Attribute Recognition (PAR) on resource-constrained embedded devices, this paper proposes the Lightweight Attribute Localization Model (LWALM). LWALM innovatively introduces CP Decomposition with Error-Preserving Correction (CPD-EPC) — the first application of this technique to layer-wise tensor compression in PAR models — effectively preserving gradient directions and attribute localization capability while drastically reducing model complexity. Experiments on the PA-100 and RAPv2 benchmarks demonstrate state-of-the-art lightweight performance: LWALM reduces model parameters by 72% and FLOPs by 68%, with an average accuracy drop of less than 1.2%. The method thus achieves an exceptional balance between inference efficiency and recognition accuracy, enabling practical deployment in edge scenarios.
📝 Abstract
Pedestrian Attribute Recognition (PAR) deals with the problem of identifying features in a pedestrian image. It has found interesting applications in person retrieval, suspect re-identification and soft biometrics. In the past few years, several Deep Neural Networks (DNNs) have been designed to solve the task; however, the developed DNNs predominantly suffer from over-parameterization and high computational complexity. These problems hinder them from being exploited in resource-constrained embedded devices with limited memory and computational capacity. By reducing a network's layers using effective compression techniques, such as tensor decomposition, neural network compression is an effective method to tackle these problems. We propose novel Lightweight Attribute Localizing Models (LWALM) for Pedestrian Attribute Recognition (PAR). LWALM is a compressed neural network obtained after effective layer-wise compression of the Attribute Localization Model (ALM) using the Canonical Polyadic Decomposition with Error Preserving Correction (CPD-EPC) algorithm.
