GordoNN is a model for speech recognition for neuromorphic cochleas based on two implementations of HOTS, the first being the zero dimentional HOTS from Germain, Kevin and Ryad, and the second the classic one dimensional HOTS.
Notice that due to memory limitation the Datasets are not loaded on the repo, therefore, in the case of the On and Off recognition task, the folders containing the data (on_aedats,off_aedats) should be positioned in the DATA/On_Off/ folder

(One of the biggest problems to deal with cochleas is to account for the different gain values of each filter)