Detect and correct anomalies on the fly in time series data. This was designed to accept the outputs of models like YOLO and OpenPose that typically perform their analysis on a frame-by frame basis. The goal of this package is to improve the output of other models by considering prior observations. In the context of multi-object detection where identification is not performed (such as YOLO), overseer can track objects by id.
- Download the YOLO checkpoint (initial-model.h5) from: https://github.com/ndaidong/yolo-person-detect
- Put the file in the yolo folder.
- Clone the OpenPose repository: https://github.com/CMU-Perceptual-Computing-Lab/openpose
- Download the OpenPose coco and/or body25 weights by running getModels (located in the models folder of the openpose repository).
- Add the path to openpose-master to the conf.py file (the openpose_repo_path variable).
Run front_end.py
By default it streams images from a webcam to YOLO.