Hearing the unheard, seeing the unseen

"Hearing the unheard, seeing the unseen" is a transformative deep learning project that I undertook over two months, from October to December 2022. In this endeavor, I developed a remarkable deep learning solution that reached 96% accuracy in classifying long-range video sequences of Indian sign language gestures into eight essential categories of emergency words.

Leveraging state-of-the-art technologies, I utilized a pre-trained Vision Transformer (ViT) for feature extraction and harnessed the power of a BiLSTM network for accurate sequence classification. This project showcased my proficiency in using key tools and frameworks such as Python, OpenCV, Katna, HuggingFace, Keras, and TensorFlow to make a meaningful impact in the field of deep learning and accessibility.