It contains all the code for AMLS2 Assessment Required packages include: scipy, numpy, pandas, math, os, tqdm, matplotlib, sklearn, tensorflow Here is the data link from one drive : https://liveuclac-my.sharepoint.com/:f:/g/personal/uceegy1_ucl_ac_uk/EtGBShtTJl5KhdwjtwXdtr4BS-cwEzWNUvGOjk3drboqOA?e=Eu6xlF If you want to generate input from scratch .wav file, you need to download raw data from my given link, you need uncomment first section in main.py, provide corrected relative path from your side to line 6, 9, 11, 13 in main.py and run the TFR_PLOT.m file If you use generated images by me, you need to download data from my given link. Provide a relative path from your side to line 18 in main.py that contains image data in main.py Densenet needs to be trained where normally take 1.5 hours. Custom CNN can be loaded from Results folder. Provide a relative path from your side to line 28 in main.py
!!! Please bear this in mind !!! This work shows successful in both kaggle and local machine, if you fail to repo it, please contact me and I can show you all the original work, thanks a lot for your understanding.