Listening-to-Sound-of-Silence-for-Speech-Denoising
Listening-to-Sound-of-Silence-for-Speech-Denoising copied to clipboard
henryxrl
Inference Question
Could you provide more information as to how to run the inference models? For example, how do you modify the code to point to the dataset? Any extra details or examples would be greatly appreciated!
If you want to use your own data, then follow the step to preprocess the data first. If you use the provided sample data, then you only need to modify the csv file so that it points to the correct wav file locations.
Thanks for your response! I'm not sure what you mean by "modify the csv file so that it points to the correct wav file locations" since, to my understanding, the csv file is only meant to contain the names of the audio files and not their directories. How does predict.py know where the data is stored if the code isn't modified and it's not given as an input? Does it have to be in the directory structure given in the example?
I have a issue about the pesq metrics. I see that you use the pypesq package to compute the pesq. But this package just can compute narrow band pesq version. The demand dataset is 16k sampling rate and the baseline model also provide the wide band pesq version. The two values will be different.
