Yugesh AV issues

Results 7 issues of


                                            Yugesh AV

Training and Preprocessing related

Hi, I tried preparing DNS Sample data for 50hours and got clean, noise, noisy folders which contains audio data by running noisyspeech_synthesizer_singleprocess.py As per the code in config/common/fullsubnet_train.toml it requires...

Does the Pretrained Model available in releases folder works with 48k sampling rate?

@haoxiangsnr Hello, FullSubNet model works with 48k sampling rate in inferencing time? Regards Yugesh

Output Spectrum has no information from 4k to 11k by using pretrained model (generator_v3)

Hi I have used the pretrained model (generator_v3) to inference my input signal with 22050 sampling rate which is having band information upto 4K. I could see that output band...

Model Size and Computation?

Hello Please let me know the model size and computation. Any other benchmarking available please post here Regards Yugesh

audio[1] = audio[1][:,:,self.L//2:-(rest+self.L//2)].contiguous() # B, 1, T IndexError: list index out of range

Input Feature to TRUNET

Hi As per the paper, 4 features must be concatenated as input to TRUNET, 1. log spectrum 2. PCEN 3. real part of demodulated phase 4. imaginary part of demodulated...

Can you please share the complete training scripts ?

It will be good for the researchers / students if you share the working version of training script