About

Souvik Ghosh is a software engineer at Fulltrack AI with a keen interest in deep learning, computer vision, and AI. With a strong foundation in programming languages such as C, C++, Java, and Python, Souvik is constantly striving to learn more and become a competitive programmer. At Fulltrack AI, Souvik has been involved in developing a range of innovative projects, including a multi-class classifier that classifies batsman hand (left/right/none) and wicket-keeper (present/absent), a shot classification model that identifies 21 different cricketing shots and whether they are lofted/grounded and aggressive/defensive, and an audio classification model that detects if there was a hit sound in the video. Souvik has also developed a shot similarity algorithm that generates a sorted list of shots based on similarity score, taking into account factors such as speed, spin, swing, release point, pitch point, and beehive points. To support these projects, Souvik has built annotation tools in PyQt and OpenCV, and has used Pytorch to build models with video sessions fetched from AWS S3 bucket. Souvik holds a Master of Computer Applications (MCA) degree from Pondicherry University, which he is due to complete in 2022, and a Bachelor of Science (BSc) degree in Computer Science from St. Xavier's College (Autonomous), Kolkata. With 1.78 years of relevant experience, Souvik's technical skillset includes software engineering, deep learning, computer vision, AWS, video classification, frontend development, CNN, OpenCV, Pytorch, Python, and Java.