Search papers, labs, and topics across Lattice.
Technische Hochschule Nürnberg Georg Simon Ohm,Germany
2
0
4
1
Turns out where you look in Wav2vec 2.0's representations *really* matters: intelligibility lives in the layers, while articulation problems hide in the time steps.
Wav2Vec 2.0, despite boosting accuracy in speech-based cognitive impairment detection, introduces significant gender and age biases, misclassifying females and younger individuals at a higher rate.
