Completed
Computational Imaging Systems
Class Central Classrooms beta
YouTube playlists curated by Class Central.
Classroom Contents
Computational Imaging Systems: From DiffuserCam to Neural Activity Tracking - Seminar 2
Automatically move to the next video in the Classroom when playback concludes
- 1 Intro
- 2 Computational Imaging Systems
- 3 Computational imaging pipeline
- 4 Lenses map points to points
- 5 Mask-based cameras multiplex
- 6 DiffuserCam: stick a scatterer on a sensor
- 7 Traditional cameras take direct measurements
- 8 Computational cameras can multiplex
- 9 DiffuserCam forward model is a convolution
- 10 Video from stills with rolling shutter
- 11 Point spread function shifts and scales with posit
- 12 Single-shot 3D is difficult
- 13 Compressed sensing to the rescue! solves under-determined problems via a sparsity prior
- 14 3D neural activity tracking
- 15 Neural activity tracking with flat DiffuserScope
- 16 Improved diffuser for low light
- 17 Keeping the objective lens is good
- 18 Resolution is more uniform
- 19 Tiny microscope version
- 20 Single focal length MLA
- 21 Multi-focal length MLA
- 22 Challenge: object-dependent resolution
- 23 Solution?: use condition number of sub-proble
- 24 Challenge #2: model mis-match
- 25 Solution #2: Local convolution model
- 26 Image reconstruction is nonlinear optimizatior
- 27 Physics-based image reconstruction
- 28 Deep learning based reconstruction
- 29 Inverse Problem Philosophies
- 30 Unrolled physics-based algorithm makes efficient ne
- 31 Physics-based learning improves speed + quali
