Class Discussion Summary (Jan 21)

Put the citations to papers discussed here:

Scribes:

Brian Schwarzmann (group 7)

Andy Vitek (group 6)

Yishu Wang(group 1)

Melanie Wong (group 3)

Write the Class Discussion Summary Here:

• What are the property of SPAD cameras which are different than conventional cameras?

Single-photon avalanche diodes are an emerging sensor technology capable of detecting individual incident photons, and capturing their time-of-arrival with high timing precision, while previous sensors were limited to single-pixel or low-resolution devices.

• What, if anything, did this paper introduce in terms of new ideas? Really think. For each idea presented, was it new, or was it known before.

SPADs directly counts the photon so it can skip DA/AD converter, which decreases noise

Not new: the author designs an algorithm that aligns and merges binary sequences captured by single-photon cameras (SPCs) into intensity images with minimal motion blur and artifacts, high signal-to-noise ratio, and high dynamic range. Some technology in the algorithm has already been discussed and researched in the previous papers, such as reducing blur while moving. The author proposed the method by utilizing a SPAD array to generate high-quality images for scenes with challenging lighting, complex geometries, high dynamic range and moving objects.

Single photons benefit computations and produce better low-light photography overall.

• Brainstorm - Given this emerging sensor (SPADs), brainstorm and then pick a new research problem or application domain? What key new thing is enabled that no one has thought of. What would be enabled if some key difficulty was resolve? A great problem is brand new, but you also have a good idea how to accomplish it.

• Since SPADs record each photon, will it consume a lot of power and RAM? Can we use it on phone?
• Enable capturing all frames in high-quality photos during a motion

• The video reconstruction method is too complicated and the computational complexity is too high. It is necessary to develop a novel algorithm to achieve temporal coherency across reconstructed frames.
• Images produced are grayscale
  • Use neural networks to provide color
  • Apply to microscopic photography
• Lidar and other distance measuring photography
  • Measures of photon travel time could have applications in physics