Presentation
Dirty Pixels: Towards End-to-end Image Processing and Perception
SessionTechnical Papers
Event Type
Technical Paper
Research & Education
Ultimate Supporter
Ultimate Attendee
Exhibitor Ultimate
Time
Location
DescriptionWe propose an end-to-end differentiable architecture that jointly performs demosaicking, denoising, deblurring, tone-mapping, and classification. We demonstrate that our model substantially improves image reconstruction and classification in low light conditions.