What is the reason why the training hardly generates any data and there is a lot of noise at the stitching points?

Guanyuansheng / TFGAN-PLC

A Temporal-Spectral Generative Adversarial Network based End-to-end Packet Loss Concealment for Wideband Speech Transmission

30 stars 11 forks source link

What is the reason why the training hardly generates any data and there is a lot of noise at the stitching points? #14

Open Janne-byti opened 10 months ago

Janne-byti commented 10 months ago

Noisy data is randomly lost at 320 points per frame, with a packet loss rate of 5% -30%. Data with different packet loss rates are mixed together, and the generated results are as follows. What is the reason for this?

PS: No modifications were made to other parts of the code

EMMALIKECODE123 commented 10 months ago

hi janne, have you solved this problem?