Pelee opencv-Image_classification

Pelee model is now supported

opencv-Image_classification

Train Image classification on CIFAR10 using tensorflow and keras.
Create a inference using opencv. OpenCV to read tensorflow graph.

Train:

python3 train.py

Reduce unwanted nodes:

python3 optimize_for_inference.py —input model1.pb —output opt_model1.pb —input_names input_1 —output_names predictions/Softmax

Convert to .pbtxt

python3 to_pbtxt.py

remove nodes with names flatten/Shape, flatten/strided_slice, flatten/Prod, flatten/stack.

Replace a node

node { name: “flatten/Reshape” op: “Reshape” input: “block5_pool/MaxPool” input: “flatten/stack” }

by

node { name: “flatten/Reshape” op: “Flatten” input: “block5_pool/MaxPool” }

Test

python3 test.py