A GUI for YOLOv5, support all the 11 inference formats that YOLOv5 supports.
The detect program is based on YOLOv5 v6.1.
Run the detect6.py
to start the GUI.
- Compared with other YOLOv5's GUI, this one has a unique advantage, this GUI can support all the 11 inference formats that YOLOv5 supports. That will be great for embeded device deploy with GUI.
Format | Model |
---|---|
PyTorch | yolov5s. pt |
TorchScript | yolov5s.torchscript |
ONNX | yolov5s.onnx |
OpenVINO | yolov5s_openvino_model/ |
TensorRT | yolov5s.engine |
CoreML | yolov5s.mlmodel |
TensorFlow SavedModel | yolov5s_saved_model/ |
TensorFlow GraphDef | yolov5s.pb |
TensorFlow Lite | yolov5s.tflite |
TensorFlow Edge TPU | yolov5s_edgetpu.tflite |
TensorFlow.js | yolov5s_web_model/ |
PaddlePaddle | yolov5s_paddle_model/ |
- This GUI can support all the inference source when using any one in the 11 inference formats. Support source:
-
source 0 # webcam img.jpg # image vid.mp4 # video path/ # directory path/*.jpg # glob 'https://youtu.be/Zgi9g1ksQHc' # YouTube 'rtsp://example.com/media.mp4' # RTSP, RTMP, HTTP stream