This repository is just a simple example of how to run an ONNX classification model trained with FashionMNIST using OpenCV
The model was trained with LeNet5 in order to be fast and also have great accuracy due to the low resolution of the training dataset
Download the model
(Problems with Openvino 2021.1) Download MO version for InferenceEngine and ForwardAsync model
To use this repository please follow the steps below:
cd simple-fashion-classifier
python3.7 -m pip install -r requirements
python3.7 main.py path/to/model/file path/to/image_or_image_folder
to run with InferenceEngine use:
python3.7 main.py path/to/model/file path/to/image_or_image_folder inference_engine
To convert the FashionMnist csv files into images create a folder called data and use:
python3.7 scripts/fashion_mnist_from_csv.py