Cascade is a robust and efficient image classification solution built on the ResNet architecture. Leveraging the power of the Microsoft Cognitive Toolkit (CNTK), Cascade makes it easy to train, evaluate, and implement deep learning models for your image classification tasks.
Note: Cascade offers an interface primarily in Russian to cater to our native users. However, all the inline code documentation is provided in English.
- ResNet Architecture: Utilizes the power of the deep residual network (ResNet) architecture for high-performance image classification.
- Training and Testing: Easy-to-use interface for training models on your data and evaluating their performance.
- User Interface Control: Built-in functions to enable or disable UI elements during training for a smoother user experience.
- Model Saving: Automatically saves models after each training epoch, allowing you to choose the best model for your task.
Clone this repository to your local machine using:
git clone https://github.com/PrimitivesLTD/Cascade.git
Then open the solution in Visual Studio and build x64 solution (tested in VS 2022)