This repository contains a project based on American Sign Language detection using template matching. Link to the drive containing the dataset captured and used: https://drive.google.com/drive/folders/1ocKtoNyEMHJC36oHiU6UhD0JIjiLqV1P?usp=sharing

Note: The 'cropped_captured_templates' folder is the one that is being used for the experimentations. The 'test_cases' folder contains 5 frames captured at random of each alphabet in a constricted environment(meaning results could vary in a different/more dynamic environment). The 'detection_session' is a live session that has been recorded of a user being able to capture the letters that they are willing and are able to save them in a text file named: 'detected_letters'. In the example above, the 'detected_letters' contain the letters 'SAUNAK' (my name). If anybody would like to try it out for themselves, they may incorporate the dataset stored in 'cropped_captured_templates' in their own python environment and use the 'Demo' python notebook.