video-object-detection

  OpenMMLab Video Perception Toolbox. It supports Video Object Detection (VID), Multiple Object Tracking (MOT), Single Object Tracking (SOT), Video Instance Segmentation (VIS) with a unified framework.

翻译 - OpenMMLab视频感知工具箱。它通过统一的框架支持单对象跟踪（SOT），多对象跟踪（MOT），视频对象检测（VID）。

  Out-of-the-box code and models for CMU's object detection and tracking system for multi-camera surveillance videos. Speed optimized Faster-RCNN model. Tensorflow based. Also supports EfficientDet. WAC...

翻译 - CMU用于监视视频的目标检测和跟踪系统的现成代码和模型。速度优化的Faster-RCNN模型。基于Tensorflow还支持EfficientDet。 WACVW'20