Deep Learning Library / Computer Vision Library
3.6k 2026-04-18
ZhaoJ9014/face.evoLVe
A high-performance, comprehensive face recognition library built on PaddlePaddle and PyTorch.
Core Features
Comprehensive face analytics (detection, landmark localization, affine transformation).
Robust data processing capabilities including augmentation and normalization.
Dual framework support: optimized for both PaddlePaddle and PyTorch.
Advanced distributed training for large-scale face recognition.
Provides feature extraction APIs from pre-trained models.
Detailed Introduction
face.evoLVe is a high-performance, comprehensive library dedicated to face-related analytics and applications, built upon the robust deep learning frameworks of PaddlePaddle and PyTorch. It offers a full suite of tools for face alignment, data processing, and efficient training/validation. A key innovation is its elegant distributed training schema, designed to overcome limitations in large-scale face recognition by supporting both backbone and head layers across multiple GPUs. This project empowers researchers and developers to build advanced, scalable face recognition systems, pushing the boundaries of computer vision applications.