Spiking-timing dependent plasticity (STDP) is a biologically observed rule of synaptic strengthening and weakening. This model implements an abstract, feed-forward STDP network which works with binocular inputs. The paper uses the Hunter-Hibbard and KITTI datasets. I encourage readers and users to try other datasets.
The code is currently hosted on ModelDB, and a git repo will be made available shortly. I hope it will be useful to both biologists and AI researchers. Please let me know if you use these ideas for any interesting (new or not) insights and applications.
Reference: Chauhan, T., Masquelier, T., Montlibert, A., & Cottereau, B. (2018). Emergence of binocular disparity selectivity through Hebbian learning. The Journal of Neuroscience, 38(44), 9563–9578. [doi]