ASL Recognition with Hidden Markov Models
AI system leveraging Hidden Markov Models (HMMs) to recognize American Sign Language gestures from video input with high confidence.
Designed and implemented an AI system that leverages Hidden Markov Models (HMMs) to recognize American Sign Language (ASL) gestures from video input.
Approach
- Extracted hand-position features from video sequences and trained HMMs on a labelled dataset of ASL signs
- Achieved accurate recognition across a vocabulary of gestures with high confidence levels
- Applied probabilistic sequence modelling to capture the temporal dynamics of sign language
Tools & Techniques
- Language: Python
- Libraries: hmmlearn, NumPy
- Course: Artificial Intelligence (CS 6601), Georgia Tech
- Period: January – May 2024