7 ms·

Wow, I'm calling a hoax. They produce images of dubious value, and for whatever reason, contaminate the results by mutating them with the Deep Dream stylization

Wow, I'm calling a hoax. They produce images of dubious value, and for whatever reason, contaminate the results by mutating them with the Deep Dream stylization. This is as bad as the dead salmon results. EDIT: You can flag all you want, but look at this playback of the reconstruction: https://video.twimg.com/tweet_video/DSrWxhSVQAE2Xix.mp4 https://video.twimg.com/tweet_video/DSrWxhSVQAE2Xix.mp4 There's nothing useful happening here.

LeoPanthera 8y ago
That's exactly what they say they're doing. That's not "contamination" - that's the whole point.
- dreepDeam 8y ago
  And that's exactly why this is garbage research. Point deep dream at an interference pattern of pure noise, after training it on a set of target images. Don't be surprised when it "finds" the image, pulled from raw static that contained no signal.
  - emileokada 8y ago
    Not convinced this is as junk as you claim it to be. From the abstract: > While our model was solely trained with natural images, our method successfully generalized the reconstruction to artificial shapes, indicating that our model indeed reconstructs or generates images from brain activity, not simply matches to exemplars. If I am reading that correctly it is reconstructing images that the subjects had not seen before.
sdinsn 8y ago
The code and data is here: https://github.com/KamitaniLab/DeepImageReconstruction https://github.com/KamitaniLab/DeepImageReconstruction Not a hoax. Your comparison to Deep Dream is quite apt, for good reason that you clearly don't realize- Deep Dream is an attempt to "enhance patterns in images via algorithmic pareidolia"- Pareidolia being a "psychological phenomenon in which the mind responds to a stimulus". So yes, this project is affected by a real life "deep dream".
- dreepDeam 8y ago
  Oh, a github repo suddenly validates scientific claims. The presence of an open repo really doesn't bolster the claims of this project. As if to say "go ahead, try it yourself. you can replicate your own results." That doesn't improve the premise of this junk science, since the images are fabricated, and poor quality besides.
bgie 8y ago
Not my field of expertise but on page 15 they claim " In our reconstruction analysis, we used a pre-trained DGN which was provided by Dosovitskiy & Brox" So if the deep dream generator is not trained on their actual (limited) training data but instead on a massive dataset of 'all photos that are real", I would say the study is valid.
teilo 8y ago
Don't feed trolls with throwaway accounts.

LeoPanthera 8y ago

That's exactly what they say they're doing. That's not "contamination" - that's the whole point.

dreepDeam 8y ago
And that's exactly why this is garbage research. Point deep dream at an interference pattern of pure noise, after training it on a set of target images. Don't be surprised when it "finds" the image, pulled from raw static that contained no signal.
- emileokada 8y ago
  Not convinced this is as junk as you claim it to be. From the abstract: > While our model was solely trained with natural images, our method successfully generalized the reconstruction to artificial shapes, indicating that our model indeed reconstructs or generates images from brain activity, not simply matches to exemplars. If I am reading that correctly it is reconstructing images that the subjects had not seen before.

sdinsn 8y ago

The code and data is here: https://github.com/KamitaniLab/DeepImageReconstruction https://github.com/KamitaniLab/DeepImageReconstruction Not a hoax. Your comparison to Deep Dream is quite apt, for good reason that you clearly don't realize- Deep Dream is an attempt to "enhance patterns in images via algorithmic pareidolia"- Pareidolia being a "psychological phenomenon in which the mind responds to a stimulus". So yes, this project is affected by a real life "deep dream".

dreepDeam 8y ago
Oh, a github repo suddenly validates scientific claims. The presence of an open repo really doesn't bolster the claims of this project. As if to say "go ahead, try it yourself. you can replicate your own results." That doesn't improve the premise of this junk science, since the images are fabricated, and poor quality besides.

bgie 8y ago

Not my field of expertise but on page 15 they claim " In our reconstruction analysis, we used a pre-trained DGN which was provided by Dosovitskiy & Brox" So if the deep dream generator is not trained on their actual (limited) training data but instead on a massive dataset of 'all photos that are real", I would say the study is valid.

teilo 8y ago

Don't feed trolls with throwaway accounts.