BioengineeringbioRxiv
Optimization of Neuroprosthetic Vision via End-to-end Deep Reinforcement Learning
B. Kücükoglu, B. Rueckauer, N. Ahmad…·M. van Gerven
This paper uses deep reinforcement learning to optimize stimulation patterns for neuroprosthetic vision, showing improved performance compared to fixed feature extractors in complex environments.
DOI 10.1142/s0129065722500526