English | δΈζ
AlexNet PetClassification is a simple image classifier based on AleNet and MindSpore to recognize cats and dogs. This project is a course design for the "ModelArts Machine Learning Practice" course of Chongqing University of Posts and Telecommunications in the spring semester of 2023.
According to the system and hardware environment, install the related dependencies of Mind Spore. For details, please refer to Mind Spore Installation Guide.
Download the open source cat and dog dataset on the Kaggle platform. For details, please refer to Cat and Dog Dataset Download.
Preprocessing can be performed in pre_process.py
to process images that do not meet our requirements.
Among them we need to rename the dataset we downloaded to cat_dog.zip
.
python pre_process.py
python train.py
After training, the accuracy of the back-end model after training will be generated.
This project is under the Apache License 2.0. See the LICENSE file for the full license text.