CONF CoppietersdeGibson_INTERSPEECH_2022/IDIAP Low-Level Physiological Implications of End-to-End Learning for Speech Recognition Coppieters de Gibson, Louise Garner, Philip N. cochlear models end-to-end architectures filterbanks SincNet speech recognition EXTERNAL https://publications.idiap.ch/attachments/papers/2022/CoppietersdeGibson_INTERSPEECH_2022.pdf PUBLIC Proc. Interspeech 2022 2022 749--753 2308-457X 10.21437/Interspeech.2022-10093 doi Current speech recognition architectures perform very well from the point of view of machine learning, hence user interaction. This suggests that they are emulating the human biological system well. We investigate whether the inference can be inverted to provide insights into that biological system; in particular the hearing mechanism. Using SincNet, we confirm that end-to-end systems do learn well known filterbank structures. However, we also show that wider band-width filters are important in the learned structure. Whilst some benefits can be gained by initialising both narrow and wide-band filters, physiological constraints suggest that such filters arise in mid-brain rather than the cochlea. We show that standard machine learning architectures must be modified to allow this process to be emulated neurally.