eejr

4 ms·

1.
Show HN: Openpilot-tools(medium.com)
9 points by eejr 8y ago|0 comments
2.
Neural Enhance – Super Resolution for images using deep learning(github.com)
221 points by eejr 10y ago|51 comments
3.
by eejr 10y ago
think open source + behind the scenes. the plan is to share code, papers, science, thoughts in the making, research, etc. but all around well defined projects to be pragmatic. if you already like projects like Keras, this may be a nice plac
4.
by eejr 10y ago
yeah, its from less than an hour ago :D We will add the welcome messages and everything else. But please, feel free to introduce yourself in the project that you think is interesting and get a conversation started.
5.
by eejr 10y ago
it was proposed by François Chollet (author of Keras) https://twitter.com/fchollet/status/788462595330105344
6.
by eejr 10y ago
this is supposed to be used in the data processing step. you load your image from jpeg or your video using ffmpeg, enhance the images and then pass it to the next step where color rendering is done. you can do that in the browser or mobile
7.
by eejr 10y ago
;) sshh! don't say that too loud yet. but, remember our names...
8.
by eejr 10y ago
yeah, no AI. Its low level computer vision. There is no implicit understanding of the scene to enhance it here. We show the neural nets several examples of low and high quality images it learns a function that makes the low quality looks mo
9.
by eejr 10y ago
yup, to certain point! there are information theoretic limits though. You can fill in information, but there will be biases to a certain point. in this case defined by the dataset. if the "enhance" is too strong, we should be care
10.
by eejr 10y ago
Hi, sorry for the poor documentation. But please, check out the paper the person above mentioned. We worked hard to make it as accessible as possible. Also, the core of our code is actually here: https://github.com/EderSanta