Skip to content

Post Hoc Denoising

DrCoffey edited this page Jul 26, 2021 · 9 revisions

Detection in DeepSqueak v3 produces generally produces far fewer false positives than v2, so a default post-hoc denoising network has not been included. A new post-hoc denoising network may be trained for recordings with particularly bad noise issue.

False positives may be automatically identified and rejected using a post hoc neural network. The network must be located in "DeepSqueak\Denoising Networks\CleaningNet.mat".

To use the post hoc denoiser:

  1. Select "Tools > Automatic Review > Post Hoc Denoising"

  2. In the list box, select the detection files to denoise. All noise events found will be classified as "Noise" and rejected.

To train a post hoc denoiser:

  1. Select "Tools > Network Training > Train Post Hoc Denoiser"

  2. In the list box, select the detection files to to use for training.

    • Calls labeled as "Noise" are used as negative training sample.

    • Accepted calls not labeled as "Noise" are used as positive training samples.

  1. When training is finished, the new network will be automatically saved as "DeepSqueak\Denoising Networks\CleaningNet.mat".

    - Training will overwrite the older network. It is wise to create a backup of the old network.