Search papers, labs, and topics across Lattice.
MILES Team, LAMSADE, Universit茅 Paris Dauphine-PSL, NLP team, GETALP Team, Universit茅 Grenoble Alpes
2
0
2
Finetuning speech recognition models can either amplify or erase critical speaker group information, depending on the training focus and fairness interventions.
Bias against certain speaker groups is embedded in self-supervised speech models from the very first layers, complicating efforts to achieve fairness in speech recognition tasks.
