pytorch>> face-attribute-prediction>> 返回
项目作者： d-li14

项目描述 ：
Face Attribute Prediction on CelebA benchmark with PyTorch Implementation
高级语言： Python
项目地址: git://github.com/d-li14/face-attribute-prediction.git
创建时间： 2019-02-01T08:46:52Z
项目社区：https://github.com/d-li14/face-attribute-prediction
开源协议：MIT License
下载


face-attribute-prediction

Face Attribute Prediction on CelebA benchmark with PyTorch Implemantation, heavily borrowed from my MobileNetV2 implementation.

Dependencies

  • Anaconda3 (Python 3.6+, with Numpy etc.)
  • PyTorch 0.4+
  • tensorboard, tensorboardX

Dataset

CelebA dataset is a large-scale face dataset with attribute-based annotations. Cropped and aligned face regions are utilized as the training source. For the pre-processed data and specific split, please feel free to contact me: d-li14@mails.tsighua.edu.cn

Features

  • Both ResNet and MobileNet as the backbone for scalability
  • Each of the 40 annotated attributes predicted with multi-head networks
  • Achieve ~92% average accuracy, comparative to state-of-the-art
  • Fast convergence (5~10 epochs) through finetuning the ImageNet pre-trained models