https://github.com/iwillstealyouroreos/Nerd-Detection-Software
Using OpenCV's computer vision library, I developed a program to detect whether someone is wearing glasses. To achieve this, I began by utilizing the Haar facial recognition dataset to train a classifier on the general characteristics of a face. This included various skin tones and facial features of individuals without glasses. I then applied a classic conditional probability approach to identify faces with glasses. This in turn allowed the model to classify faces with glasses; determine if they were a nerd or not.
This project showcases my technical skills in computer vision and machine learning, including working with datasets, training models, and using probability-based classification. It highlights my ability to use OpenCV for practical applications and design models that account for diverse inputs. As you can see in the video, my face is still detected even under purple lighting that accentuates my darker toned features. Overall, this reflects my ability to develop solutions that echo diverse planning.
