Other resources

Other resources shared with the public are available in this page. Feel free to contact us for further questions, or to share your interest in specific aspects of our research.

RTL code for custom blocks used in our chips
As an example, the VHDL code of the frequency-domain feature extractor of our work on ultra-low power voice activity detection (VAD) systems can be found here.
The VHDL code includes both an FFT256 and a FFT64 ultra-low power core validated on silicon (see related publications), and a feature extractor (FE) based on Mel-spaced bands.
Some descriptive material can be found here. Details on the architecture are available in our related publications (J. H. Teo, K. Ali, M. Alioto, “Voice Activity Detection with >83% Accuracy under SNR down to -3dB at 1.19µW and 0.07mm2 in 40nm,” in Proc. of IEEE ASSCC 2020, Nov. 2020).

For use in publications and presentations please cite this data collection as follows:
J. H. Teo, M. Alioto, RTL description of "Voice Activity Detection with >83% Accuracy under SNR down to -3dB at 1.19µW and 0.07mm2 in 40nm," [Online]. Available: http://www.green-ic.org/other-resources

Other examples will be regularly posted in this page, for the benefit of our community.