Staff AI Scientist - Computer Vision

Cutting-Edge, Hands-On Research Invent and develop next-generation AI-driven image and video processing technologies, with a particular focus on image/video capture solutions using emerging sensors, camera signal processing, and 3D spatial capture. Combine traditional image processing techniques with advanced AI algorithms to address challenges in image delivery, analysis, and content creation. Guide and mentor researchers specializing in image processing, computer graphics, and content enhancement. Act as a coach and mentor to junior researchers, helping them grow into independent contributors within a corporate research environment. Foster a dynamic, transparent, results-driven, and innovative working culture. Ph.D. in Physics, Electrical Engineering, Mathematics, or Computer Science, plus 4+ years of corporate research experience. Profound expertise in image sensors, lenses, camera signal processing, spectral processing, 3D modeling, geometry, and calibration. Deep theoretical and practical understanding of state-of-the-art AI technologies, including knowledge of network architectures, deployment modes, data augmentation, and performance analysis. Generative models (diffusion, autoregressive, etc.) Self-supervised learning, contrastive learning, and autoencoders Applications such as source separation, text-to-speech, music synthesis, image segmentation, and language models. Familiarity with gaming environments, SoC architectures, embedded systems, and GPU/CPU implementations. Proven ability to invent, patent, and productize novel image processing and rendering technologies. Strong publication record in leading conferences and journals (e.g., NeurIPS, ICLR, ICML). Expertise in C/C++ and Python. Demonstrated leadership in managing and mentoring researchers. Passion for cross-continental teamwork and collaboration. Strong communication, collaboration, and presentation skills in English. A forward-thinking innovator with the ability to envision and create impactful, commercially viable technologies.