4 ms·

> "That all sounds like science fiction, that can't happen in the real world." Tell them that the mobile phone they have in their pocket also seems like scienc

by imron 11y ago

> "That all sounds like science fiction, that can't happen in the real world." Tell them that the mobile phone they have in their pocket also seems like science fiction - and it was 20-30 years ago. Now it is science fact. Facial recognition is getting the same way. If anything, based on all the shoddy movies that 'zoom in, enhance' with photos, the public should be primed to accept facial recognition quite easily.

duderific 11y ago
Ha, when watching those shows, I am always the nerd telling my wife how it's impossible to zoom in and enhance, because you can't create more information than existed when the photo was taken. I'm not sure that she gets it.
- meric 11y ago
  If you take a blurry photo of a license plate - it can be possible to enhance it (e.g. using sharpen) to reveal the license plate number - just because your naked eyes can't extract that information from a blurry photo doesn't mean the blurry photo doesn't already have enough information to encode the number precisely.
  - yorak 11y ago
    There is also superresolution methods that in practice make the "enhance"-feature possible. The information from the same objects in successive frames, eg. from a security camera feed, are composited into a single frame of higher resolution than the originals. This can even be done for a single image (see. Glasner et al. "Super-resolution from a single image." Computer Vision, 2009 ICCV 2009. IEEE.). If machine learning is brought into play it gets even more interesting. See eg. waifu2x, the anime video upscaler based on Deep Convolutional Neural Networks. Basically it is a neural net, that has been shown A LOT of anime, and it can therefore creatively(?) fill in the created pixels when upscaling. I have no idea how advanced this kind of ML boosted "enhance" technology can get in the future, and it might even get a little scary: imagine an observer with the experience of million lives spent looking blurry images in poor lighting. It might get surprisingly good at that...
shabbyrobe 11y ago
I know it seems crazy, it seems crazy to me too, but this is what I encounter time and time again. No matter how carefully you explain it, it seems too unlikely to them. Once the explanation passes a certain threshhold of complexity (for which the bar is frighteningly low), they tune out and shut down.
- Estragon 11y ago
  Maybe what's needed more than an elevator pitch is an open-source app using facial recognition to creep people out without abusing them.
  - sdoering 11y ago
    Why not abuse them. If they let themselves be led to the slaughterhouse that easily and voluntarily maybe they have to feel the pain of being abused to learn? I get more cynical by the day, the more I read/hear such accounts. I am/was an idealist, but I catch myself every so often to think, that these people do not deserve anything else. Everything is clearly visible - that possible paths are not that hard to understand. The consequences can be dire if nothing is done - non the less. They keep giving away their and everyone else's privacy in the blink of a thumb-flick on "what ever the next big app is called". Have I given up? No, I try non the less to inform people surrounding me. Do I believe in positive change? Sadly not anymore.
    - kaybe 11y ago
      I tried to discuss the dual-use nature of their research with some machine learning/pattern recognition people (I was there to discuss maybe working with them as a student). They didn't seem to ever have thought about it at all, and were not open to start now. This was quite shocking to me. I'm not sure how to deal with the issue when even the people creating the technology are like that.

by imron 11y ago

duderific 11y ago

Ha, when watching those shows, I am always the nerd telling my wife how it's impossible to zoom in and enhance, because you can't create more information than existed when the photo was taken. I'm not sure that she gets it.

meric 11y ago
If you take a blurry photo of a license plate - it can be possible to enhance it (e.g. using sharpen) to reveal the license plate number - just because your naked eyes can't extract that information from a blurry photo doesn't mean the blurry photo doesn't already have enough information to encode the number precisely.
- yorak 11y ago
  There is also superresolution methods that in practice make the "enhance"-feature possible. The information from the same objects in successive frames, eg. from a security camera feed, are composited into a single frame of higher resolution than the originals. This can even be done for a single image (see. Glasner et al. "Super-resolution from a single image." Computer Vision, 2009 ICCV 2009. IEEE.). If machine learning is brought into play it gets even more interesting. See eg. waifu2x, the anime video upscaler based on Deep Convolutional Neural Networks. Basically it is a neural net, that has been shown A LOT of anime, and it can therefore creatively(?) fill in the created pixels when upscaling. I have no idea how advanced this kind of ML boosted "enhance" technology can get in the future, and it might even get a little scary: imagine an observer with the experience of million lives spent looking blurry images in poor lighting. It might get surprisingly good at that...

shabbyrobe 11y ago

I know it seems crazy, it seems crazy to me too, but this is what I encounter time and time again. No matter how carefully you explain it, it seems too unlikely to them. Once the explanation passes a certain threshhold of complexity (for which the bar is frighteningly low), they tune out and shut down.

Estragon 11y ago
Maybe what's needed more than an elevator pitch is an open-source app using facial recognition to creep people out without abusing them.
- sdoering 11y ago
  Why not abuse them. If they let themselves be led to the slaughterhouse that easily and voluntarily maybe they have to feel the pain of being abused to learn? I get more cynical by the day, the more I read/hear such accounts. I am/was an idealist, but I catch myself every so often to think, that these people do not deserve anything else. Everything is clearly visible - that possible paths are not that hard to understand. The consequences can be dire if nothing is done - non the less. They keep giving away their and everyone else's privacy in the blink of a thumb-flick on "what ever the next big app is called". Have I given up? No, I try non the less to inform people surrounding me. Do I believe in positive change? Sadly not anymore.
  - kaybe 11y ago
    I tried to discuss the dual-use nature of their research with some machine learning/pattern recognition people (I was there to discuss maybe working with them as a student). They didn't seem to ever have thought about it at all, and were not open to start now. This was quite shocking to me. I'm not sure how to deal with the issue when even the people creating the technology are like that.