Robustness recording condition

Overall scores

CNN14

w2v2-b

hubert-b

axlstm

Overall Score

0.0% passed tests (0 passed / 2 failed).

0.0% passed tests (0 passed / 2 failed).

50.0% passed tests (1 passed / 1 failed).

0.0% passed tests (0 passed / 2 failed).

Percentage Unchanged Predictions Recording Condition

Threshold: 0.8

Data

Percentage Unchanged Predictions Recording Condition

CNN14

w2v2-b

hubert-b

axlstm

imda-nsc-read-speech-balanced-2.6.0-headset-boundary

0.49

0.77

0.82

0.52

imda-nsc-read-speech-balanced-2.6.0-headset-mobile

0.23

0.57

0.72

0.46

mean

0.36

0.67

0.77

0.49

Visualization

Difference of predictions for baseline recording condition audio and different recording condition audio. The allowed prediction difference \(\delta < 0.05\) is highlighted in green in the upper plot. The lower plot shows the distributions of the two predictions.

CNN14

w2v2-b

hubert-b

axlstm

../../../_images/visualization_imda-nsc-read-speech-balanced-2.6.0-headset-boundary11.png
../../../_images/visualization_imda-nsc-read-speech-balanced-2.6.0-headset-boundary12.png
../../../_images/visualization_imda-nsc-read-speech-balanced-2.6.0-headset-boundary13.png
../../../_images/visualization_imda-nsc-read-speech-balanced-2.6.0-headset-boundary14.png
../../../_images/visualization_imda-nsc-read-speech-balanced-2.6.0-headset-mobile11.png
../../../_images/visualization_imda-nsc-read-speech-balanced-2.6.0-headset-mobile12.png
../../../_images/visualization_imda-nsc-read-speech-balanced-2.6.0-headset-mobile13.png
../../../_images/visualization_imda-nsc-read-speech-balanced-2.6.0-headset-mobile14.png