timesler/facenet-pytorch

Implement alternative detection algorithms

Open

#74 创建于 2020年4月16日

在 GitHub 查看
 (0 评论) (2 反应) (0 负责人)Python (830 fork)batch import
help wanted

仓库指标

Star
 (3,830 star)
PR 合并指标
 (30 天内没有已合并 PR)

描述

As highlighted in #10, the current MTCNN design is appropriate for the project while a single detection method is implemented, but with multiple, it makes more sense to separate the return of bounding boxes and confidences from the cropping of images.

Should also think about returning facial landmarks to implement face alignment following detection.

Retinaface is probably the best option to start with.

贡献者指南