Sign Language Recognition using Gait Energy Image
Gait Recognition with 3D CNN. This project proposes a novel approach using 3D convolutional neural networks (3D CNN) to capture spatio-temporal features of gait sequences for robust recognition in an un-intrusive manner.
Analysis code for the manuscript 'Reliability-aware anomaly detection of dairy cow lameness from side-view gait and energy images under commercial-farm crowding', under review at Scientific Reports and not published. Mirror of the Zenodo deposit.
This project leverages state-of-the-art deep learning models—VGG-16, InceptionV1, ViT, EfficientNet-B0, and ResNet50—for gait image classification. By applying decision level fusion, feature level fusion, and hybrid fusion techniques, the project achieves enhanced accuracy, making it suitable for any image classification task.
To associate your repository with the gait-energy-image topic, visit your repo's landing page and select "manage topics."