Showing 1–1 of 1 results for author: Ribeiro, B P
-
EdgeEar: Efficient and Accurate Ear Recognition for Edge Devices
Authors:
Camile Lendering,
Bernardo Perrone Ribeiro,
Žiga Emeršič,
Peter Peer
Abstract:
Ear recognition is a contactless and unobtrusive biometric technique with applications across various domains. However, deploying high-performing ear recognition models on resource-constrained devices is challenging, limiting their applicability and widespread adoption. This paper introduces EdgeEar, a lightweight model based on a proposed hybrid CNN-transformer architecture to solve this problem.…
▽ More
Ear recognition is a contactless and unobtrusive biometric technique with applications across various domains. However, deploying high-performing ear recognition models on resource-constrained devices is challenging, limiting their applicability and widespread adoption. This paper introduces EdgeEar, a lightweight model based on a proposed hybrid CNN-transformer architecture to solve this problem. By incorporating low-rank approximations into specific linear layers, EdgeEar reduces its parameter count by a factor of 50 compared to the current state-of-the-art, bringing it below two million while maintaining competitive accuracy. Evaluation on the Unconstrained Ear Recognition Challenge (UERC2023) benchmark shows that EdgeEar achieves the lowest EER while significantly reducing computational costs. These findings demonstrate the feasibility of efficient and accurate ear recognition, which we believe will contribute to the wider adoption of ear biometrics.
△ Less
Submitted 11 February, 2025;
originally announced February 2025.
