a custom object detection with cnn and rasialblocks and ready to train with at least 100 classes
Run :
python Main.py
layers : created with 3 sort of layers you can check in Layers.py and reading configs form config.py
config : config of layers with the size of each neuron or layer type or mode
you can read the model layers in the layers.txt
run for detect :
python opencv_test.py --weights ($ONNX File ) --source (video from anywhere or web cam (0)) --size 640
