BEGIN:VCALENDAR VERSION:2.0 PRODID:-//Memento EPFL// BEGIN:VEVENT SUMMARY:IC Colloquium: Cameras for better machine learning\, and machine l earning for better cameras DTSTART:20230309T100000 DTEND:20230309T110000 DTSTAMP:20260407T045653Z UID:fe0ee45527274ec383393d43ac0f2736a909b6cdc3ce3c1afeea771f CATEGORIES:Conferences - Seminars DESCRIPTION:By: Kristina Monakhova - MIT\nIC Faculty candidate\n\nAbstract \nCameras are everywhere powering everything from self‑driving cars to m edical diagnostics and scientific discovery\, creating massive economic va lue and saving (or failing to save) human lives. But automation has fundam entally changed their purpose: people don’t look at these images\, algor ithms do. Our current cameras are optimized for the wrong goal: sharp\, hi gh‑resolution images for human consumption\, rather than information‑d ense images to be used by algorithms. Using machine learning\, we can desi gn better\, more capable imaging systems\; with better imaging systems\, w e could have more robust\, better-informed intelligent systems.\n\nIn this talk\, I will demonstrate how we can make imaging systems more capable by using physics-informed machine learning\, which combines imaging system p hysics with deep learning. First\, I will show how we can make tiny\, lens less cameras have image quality comparable to lensed cameras. Next\, I wil l show how with our algorithms\, we can push the limit of what cameras c an see in the extreme dark by an order of magnitude\, enabling photoreal istic videos of moving objects on a clear\, moonless night with no exter nal illumination (submillilux). In addition\, I will demonstrate how we ca n design cheap\, compact\, and capable computational cameras and microscop es that capture higher-dimensional information\, such as 3D and multiple w avelengths of light\, which could be useful for high-level tasks. My fut ure research agenda expands upon this to design optics and algorithms toge ther with higher level tasks in order to create the next generation of i ntelligent cameras and microscopes that are optimized for high‑level ins ights rather than images.\n\nBio\nKristina Monakhova is a postdoctoral fel low at MIT\, supported by the MIT Postdoctoral Fellowship for Engineering Excellence. She received her PhD from UC Berkeley in Electrical Engineerin g and Computer Sciences in 2022\, where she was a member of Laura Waller ’s Computational Imaging research group. Her research focuses on making more capable cameras and microscopes through the co-design of imaging sy stems and algorithms. Her research lies at the intersection of signal pr ocessing\, machine learning\, and optics. She is a recipient of the NSF Graduate Research Fellowship and the NDSEG Fellowship.\n\nMore information \n  LOCATION:BC 420 https://plan.epfl.ch/?room==BC%20420 https://epfl.zoom.us/ j/69097990761 STATUS:CONFIRMED END:VEVENT END:VCALENDAR