Open gurugubelllik opened 1 year ago
I appreciate your interest in my work and would like to extend my gratitude. However, I regret to inform you that it requires contractual agreements to publicly release the official code. Despite this restriction, I am more than willing to provide assistance in other ways. If you have questions related to the concepts, methodologies, or any other aspects of the research, I'm here to help provide clarifications. Please feel free to reach out with any inquiries you may have.
In the paper, you are passing mel spectrogram to 1D convolution. How you are passing the 2D matrix to 1D array? Are you flattening it? if mean it will loose the time information right?
Thankx, kesav
Implementation of "Learning Audio-Text Agreement for Open-vocabulary Keyword Spotting" was not found in this repo. Any other sources are there for this approach?