cansik / deep-vision-processing

Deep computer-vision algorithms for the Processing framework.
91 stars 22 forks source link

can we implement Yolov 5 and 6 with custom data weights #18

Open synersignart opened 1 year ago

synersignart commented 1 year ago

can we implement Yolov 5 and 6 custom on this too , and possibly allowing users to add their own custom trained data weights in the form if .pt file and .yaml files , id be also interested as the current deep vision processing I cant find any information of implementing custom datasets

thanks

cansik commented 1 year ago

It is already implemented to use YoloV5 with custom weights and models. Just export the trained model & weights to an onnx file (https://github.com/ultralytics/yolov5/issues/251) and load it:

YOLONetwork network = new YOLONetwork(
        null,
        new Path("path-to-your-onnx"),
        640, 640,
        true
);

network.loadLabels(new Path("path to a text file with labels"));
network.setTopK(100);

YoloV6 is not supported at the moment.