A very useful industrial solution could be the detection of people not wearing helmets in worksites and giving warnings for their safety. To demonstrate, see the output of the YOLOX model trained with this dataset using roboflow.
YOLOX is a high-performance YOLO, particularly suited for edge as it supports ONNX, TensorRT, and OpenVINO. Hardware optimization of the trained model to OpenVINO model is discussed in Solution #5.
Discussions
Become a Hackaday.io Member
Create an account to leave a comment. Already have an account? Log In.