Active vision with uncertainty-based saccade generation

[tcstewar]

There's an interesting paper here:

https://www.biorxiv.org/content/10.1101/2020.02.14.948992v1

It's using a generative RNN that's given a small foveated input from a visual image, and then generates a saccade based on where in the reconstructed image is the highest uncertainty. This might be an interesting basis for a more complex biological model of this whole process.