معرفی
Michael Hobley is a Postdoctoral Researcher at CalTech focusing on AI for ecology, working as the Head Scientist on FishEye. He is the founder of Noonien, a Machine Learning consultancy that delivers bespoke solutions to clients. His academic background includes a DPhil in Computer Vision and Machine Learning from the University of Oxford.
- DPhil, Computer Vision and Machine Learning, University of Oxford
Michael's research spans ecology, weakly-supervised computer vision, large-scale data understanding, generative AI, and engineering pipeline acceleration. He specializes in developing innovative AI methodologies for diverse applications, including optical engine analysis and neural surface reconstruction.
His recent publications highlight advancements in low-SNR video denoising (SAVeD), machine learning benchmarking for engine research (EngineBench), and class-agnostic counting techniques (MCAC, ABC Easy as 123). These works demonstrate his interdisciplinary approach to solving complex problems.
Michael's work has been presented at major conferences including ECCV, CVPR, and ICLR, and he has contributed datasets like MCAC and FSC-133. He leads projects at the intersection of AI and engineering, emphasizing practical applications in both academic and business contexts.
Current affiliations include CalTech's FishEye project and Noonien's bespoke ML solutions. He has previously collaborated with the Active Vision Laboratory at Oxford, developing core vision algorithms and datasets.
.jpg)
