Code to reproduce plots from the paper: Time-Frequency and Geometric Analysis of Task-dependent learning in Raw Waveform based Acoustic Models
Response Spectra for Speech Classification
![](https://github.com/Cross-Caps/SWAM/raw/main/Plots/speech_classification_3dplots.gif?raw=true)
Response Spectra for Male Speaker Classification
![](https://github.com/Cross-Caps/SWAM/raw/main/Plots/speaker_male_classification_3dplots.gif?raw=true)
Response Spectra for Female Speaker Classification
![](https://github.com/Cross-Caps/SWAM/raw/main/Plots/speaker_female_classification_3dplots.gif?raw=true)
- For implementation of short-time response spectra (STRS) and cumulative response spectra (CRS) maps; see Script
- For implementation of various geometric properties; see Script
- A demo on how to plot STRS and CRS maps is available in Notebook
- Dictionary properties K. Skretting and K. Engan, “Learned dictionaries for sparse image representation: properties and results,” in Wavelets and Sparsity XIV. International Society for Optics and Photonics, 2011, vol. 8138, pp. 404 – 417, SPIE
- WV Contours Samer A. Abdallah and Mark D. Plumbley, “If the independent components of natural images are edges, what are the independent components of natural sounds?,” in International Workshop on Independent Component Analysis and Blind Separation (ICA), September 2001, pp. 534–539.
Devansh Gupta devansh19160@iiitd.ac.in
Vinayak Abrol abrol@iiitd.ac.in