EnsFace: An Ensemble Method of Deep Convolutional Neural Networks with Novel Effective Loss Functions for Face Recognition
Open Access
- 1 December 2022
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 231-238
Abstract
In recent years, designing new effective Deep Convolutional Neural Network (DCNN) architectures and loss functions are two crucial trends in improving face recognition (FR) accuracy. However, building an optimal DCNN and ameliorating FR performance are still the main challenges for researchers. Thus, we first investigate and analyzes the effect of several novel effective loss functions based on softmax on DCNN with the ResNet architecture. We then propose an ensemble learning, namely EnsFace, by taking advantage of recent novel FR methods based on CosFace, ArcFace, and MagFace. EnsFace elaborates on the voting mechanism that utilizes non-optimal pre-trained models to obtain better discriminative ability for FR. To prove the performance of both the speed and accuracy of EnsFace, we carry out rigorous experiments using several popular benchmarks, including LFW, CFP-FP, and AgeDB-30, as well as two renovations of LFW: CALFW and CPLFW. The results of our experiments achieve state-of-the-art figures, which show the proposed method’s massive potential to improve FR performance. Ablation studies and overall benchmarks undeniably prove the effectiveness of our EnsFace.Keywords
This publication has 12 references indexed in Scilit:
- Deep CNN ensemble for recognition of face images2021
- Optuna2019
- VGGFace2: A Dataset for Recognising Faces across Pose and Age2018
- Additive Margin Softmax for Face VerificationIEEE Signal Processing Letters, 2018
- Deep Learning for Understanding Faces: Machines May Be Just as Good, or Better, than HumansIEEE Signal Processing Magazine, 2018
- AgeDB: The First Manually Collected, In-the-Wild Age Database2017
- Frontal to profile face verification in the wild2016
- FaceNet: A unified embedding for face recognition and clustering2015
- Going deeper with convolutions2015
- DeepFace: Closing the Gap to Human-Level Performance in Face Verification2014