I developed and deployed deep learning & ML models aimed at enhancing media content accessibility and user engagement, while mentoring junior engineers, guiding them through the complexities of machine learning projects.
My projects included:
● Sports Video Classification: Leveraged deep learning models to accurately classify sports videos, enabling more effective content organization and retrieval.
● Content Safety: Developed models to filter Not Safe For Work (NSFW) thumbnails, ensuring a safe viewing environment
● Thumbnail Quality Assessment: Designed and deployed algorithms for predicting thumbnail quality, significantly improving the visual appeal and effectiveness of video previews.
● Resolution Enhancement: Created models to improve lower-quality thumbnails, to enhance visual clarity without compromising on load times.
● Optical Character Recognition (OCR): Created an OCR pipeline specifically for books in various Indian vernacular languages, significantly improving text recognition accuracy and facilitating the digitalization of valuable literary content.
● PDF to EPUB Conversion: Developed a robust pipeline to convert PDF documents into the EPUB format, streamlining content accessibility across different digital platforms.
● Video Recommendation System: Implemented a recommendation engine using frequent pattern mining, information retrieval algorithms, and clustering of thumbnail and text embeddings, delivering personalized content recommendations to enhance user engagement.
● Scalability: Successfully scaled the recommender systems to accommodate over 1 million users, optimizing performance and ensuring a seamless user experience across the board.
These initiatives not only highlighted my technical expertise in developing and deploying AI and machine learning solutions but also underscored my leadership in mentoring emerging talent.