Work

AMS-Net

Year: 2026

Type: software

Abstract: Code for *AMS-Net: Lightweight Self-Attention Enhanced MobileNetV3 for Fine-Grained Image Classification on Edge Devices* (S. Wang, J. Shen, J. Yang). AMS-Net adds three Lightweight Spatial Self-Attention Modules (LSSAM) to MobileNetV3-Small, after blocks 5, 8 and 11, and trains the network with a three-phase dynamic transfer learning (DTL) schedule.

Source: Zenodo (CERN European Organization for Nuclear Research)

DOI: 10.5281/zenodo.23299140

Authors: Wang, Shitong, Shen, Jiaquan, Yang, Jiaxi

Institutions: Universiti Sains Malaysia, Luoyang Normal University, Henan University of Science and Technology

Language: English

Topic: Advanced Neural Network Applications

Subfield: Computer Vision and Pattern Recognition

Field: Computer Science

Domain: Physical Sciences

Open Access status: green

OpenAlex ID: W7221549437

Cites:

Cited by:

Related to: 10