alkalait

5 ms·

1.
State of AI for Earth Observation (preprint)(sa.catapult.org.uk)
5 points by alkalait 4y ago|1 comments
2.
by alkalait 4y ago
Hello. I'd like to share here a potentially valuable resource for those looking to understand how AI is transforming remote sensing, or get into the field of Earth Observation. (See also related Twitter thread: https://twitt
3.
by alkalait 6y ago
Last week I published this article on Amnesty International's blog, co-authored with Amnesty's human rights researchers. This is an example of AI helping to protect Human Rights. We discuss new results on satellite imagery of the
4.
Using AI to scale up human rights research: a case study on Darfur(citizenevidence.org)
1 points by alkalait 6y ago|1 comments
5.
Multi-frame super-resolution by recursive fusion(github.com)
2 points by alkalait 7y ago|0 comments
6.
Computer, Enhance Please(elementai.com)
2 points by alkalait 7y ago|0 comments
7.
by alkalait 10y ago
Getting the game state data means deciding a-priori what features the AI should learn on. The whole point of the deep learning paradigm is to allow a machine to learn such features that enable good prediction, visualization, generation (aka
8.
by alkalait 11y ago
I'm a bit familiar with PyMC, but all it seems to do is Gibbs sampling, which mixes horribly compared to HMC. How easy would the transition to PyStan be?
9.
by alkalait 11y ago
That's because most were toy examples dealing with discrete distributions in a few dimensions. Granted, these are mathematically easier to deal with, but not representative of real-world scenarios.
10.
by alkalait 11y ago
I enjoyed MMDS, but the lectures given by Ulman were painful to sit through. The man could simply not deliver a single sentence not read verbatim from a screen.