Basically I am thinking to get the audio data to the FPGA through PDM mic and then generate a MFCC spectogram from this data and pass it through a deep learning model , still in FPGA.
The choice of FPGA is to exploit the parallelism and the speed of it , so is this feasible,
Are there any resources from where I can begin learning
Thanks in Advance.

Tried to implement AI audio processing in FPGA.