Descriptive Camera: A camera that prints a description, not an image
mattrichardson.com
Descriptive Camera: A camera that prints a description, not an image
1–10 of 83 posts
Re: Descriptive Camera: A camera that prints a description, not an image
#2Like reverse Dwarf Fortress?
Re: Descriptive Camera: A camera that prints a description, not an image
#3Amazing - the meta concepts behind the "Descriptive Camera" are limited only by the imagination, the resolution of the sensing devices, the ubiquity/bandwidth of the data pipe, and the ability to manage/train the backend human workforce and maintain a consistent level of quality. I'm pretty certain, that within 3-5 years, a billion dollar company will emerge using the basic concepts captured by the "descriptive camera." That is, a remote sensing device feeding data into a human-backed work-management queue, and returning some type of structured higher-level processing of the remote-sensing device's data in close to real-time.
Just off the top of my head:
o Flora/Fauna identification
o Scene analysis (Car Crashes/Building Wreckages)
o Medical analysis
o Substance analysis
o Structural integrity review
o Online Translators. Today.
Re: Descriptive Camera: A camera that prints a description, not an image
#4This is brilliant. It reminds me of Sascha Pohflepp's Blinks & Buttons project, which swaps your photo with someone else's taken at the same time. There's even an iPhone app.
Re: Descriptive Camera: A camera that prints a description, not an image
#5Amazing - the meta concepts behind the "Descriptive Camera" are limited only by the imagination, the resolution of the sensing devices, the ubiquity/bandwidth of the data pipe, and the ability to manage/train the backend human workforce and maintain a consistent level of quality. I'm pretty certain, that within 3-5 years, a billion dollar company will emerge using the basic concepts captured by the "descriptive camer…
Medical analysis
I don't think I would rely on a description from a Mechanical Turk human for describing what they see in a pictue.
Is that a tibia or a fibula?!?
Re: Descriptive Camera: A camera that prints a description, not an image
#6Amazing - the meta concepts behind the "Descriptive Camera" are limited only by the imagination, the resolution of the sensing devices, the ubiquity/bandwidth of the data pipe, and the ability to manage/train the backend human workforce and maintain a consistent level of quality. I'm pretty certain, that within 3-5 years, a billion dollar company will emerge using the basic concepts captured by the "descriptive camer…
Medical analysis I don't think I would rely on a description from a Mechanical Turk human for describing what they see in a pictue. Is that a tibia or a fibula?!?
Right - you don't send to Mechanical Turk. You instead send it to a highly trained backend workforce, presumably highly skilled in the vertical that you are doing analysis on. And, it's not just a picture, but video/sound/other sensing. Who knows what the next advances in remote telemetry will be.
Re: Descriptive Camera: A camera that prints a description, not an image
#7I was hoping that the description would be derived algorithmically rather than via human intervention. I guess that really was a bit optimistic though :(
Re: Descriptive Camera: A camera that prints a description, not an image
#8Add speech synthesis and this would be an awesome camera for the blind.
Re: Descriptive Camera: A camera that prints a description, not an image
#9It is a cute idea but unfortunately (or fortunately?) this particular implementation of the camera comes with (a) opinions, (b) typos, and (c) probably trolls, if they know that their text will automatically be accepted and printed out.
I guess these factors could probably be reduced with more money? I don't mean paying more for the job: I mean using the API to ask a second human to verify the results of the first human and/or fix typos.
Re: Descriptive Camera: A camera that prints a description, not an image
#10It is a cute idea but unfortunately (or fortunately?) this particular implementation of the camera comes with (a) opinions, (b) typos, and (c) probably trolls, if they know that their text will automatically be accepted and printed out. I guess these factors could probably be reduced with more money? I don't mean paying more for the job: I mean using the API to ask a second human to verify the results of the first hu…
I think you're missing the point of it.