Gastonia, North Carolina, United States
Subhabrata Bhattacharya is an Advanced R&D Director at Adeia with over 20 years of experience in computer vision, machine learning, and computational imaging, where he spearheads CV/ML intellectual property development at the intersection of cutting-edge research and real-world deployment. His inventive work has yielded a growing portfolio of patent applications in emerging areas such as multimodal LLM fingerprinting, event-triggered sensor fusion, and federated learning for edge devices, reflecting a career-long commitment to translating deep technical insight into protectable, high-value innovation. He has held key technical roles at some of the world's most influential technology organizations, including Netflix, Ericsson, Adobe, Siemens, Microsoft Research, IBM, and Infosys, consistently delivering impactful systems at scale across industries. An accomplished researcher, he has published at premier venues including IEEE CVPR, ACM Multimedia, and MICCAI, and holds granted patents spanning domains from endoscopic image classification and visual effects processing to spatial anomaly detection. He earned his Ph.D. in Computer Engineering from the University of Central Florida, and his contributions have been recognized with a Best Paper Award at ICMR 2014, Grand Challenge prizes at ACM Multimedia 2013 and 2014, and two Netflix Hackday wins in the Studio category. He is an active member of the research community, having served as an invited speaker at industry and academic venues and as a program committee member for IEEE CVPR, ACM MM, and ECCV.
- Serving as Director of Advanced R&D at Adeia, leading initiatives in media space intellectual property development, focusing on innovative technologies such as federated learning, vision language models, and Gaussian splatting. - Spearheading the prototyping of cutting-edge solutions, driving advancements in AI-driven media applications while emphasizing human-centered design principles. - Actively drafting invention disclosures and filing patents, securing intellectual property for novel technologies developed within Adeia’s R&D framework.
- Leading end-to-end design and deployment of a browser-based AI pipeline enabling human-AI interaction through real-time hand gesture recognition, distraction detection (gaze recognition and facial expression analysis), and improving feedback loops for AI agents. - Pioneering cutting-edge AI solutions by leveraging advanced cloud infrastructure to build scalable, user-centric products that enhance human-computer interaction. - Developed and launched an empathetic AI avatar generation system integrating Stable Diffusion models with custom-trained language models and in-house vector databases, delivering personalized and emotionally adaptive AI-human interactions. - Demonstrated expertise in integrating multimodal Language Models (CLIP, BLLP) to create innovative, high-impact solutions for real-world applications.
- Led a team of researchers to reanimate mouth movements for online and offline dubbing etc. through a combination of algorithms from Generative models and Neural Radiance Fields - Responsibilities included: Model training/optimization, Product development, Technical evangelism
- Automated Vfx workflows for Rig Removal, Matting, and Rotoscoping using state-of-the-art video inpainting, segmentation, and contour tracking algorithms \textbf{(saved per title avg. VFX cost by 35% - Designed and implemented high-level multimedia content analysis feature extraction algorithms to groups within Netflix to improve specific tasks e.g. recommendation systems (elevated avg. viewership by 6%) - Spearheaded the effort behind the automatic creation of artwork (e.g. Billboard, Story Art, etc.) for Netflix originals using in-stream frames (reduced 70% artist effort) - Developed algorithmic solutions across content engineering teams to automate early-stage inspection including detection of dead/stuck pixels (99.2% accuracy), alignment of subtitles, and introduction of age-appropriate rating symbols (99.6% accuracy)
- Conceived innovative solutions pertaining to consumer video space, specifically video summarization, perceptive quality improvement of videos - Contributed to the Adobe After Effects engineering team towards integrating various motion graphics algorithms such as camera lens blur, motion blur