American Sign Language Recognition

Built and evaluated a CNN for static ASL gesture recognition against a Random Forest baseline using the Sign Language MNIST dataset, achieving 99.26% accuracy.

Source code

Baseline Accuracy81.64%

CNN Accuracy99.26%

Explainability ToolsSHAP & Grad-CAM

PythonPyTorchCNNsGrad-CAMSHAP

Built and evaluated a CNN for static ASL gesture recognition against a Random Forest baseline using the Sign Language MNIST dataset.

Performed hyperparameter tuning, data augmentation, and explainability analysis with SHAP and Grad-CAM.

Improved accuracy from 81.64% as the baseline to 99.26% with the CNN model.