5 papers · 1 filter
ValSub: Subsampling Validation Data to Mitigate Forgetting during ASR Personalization
Haaris Mehmood, Karthikeyan Saravanan, Pablo Peso Parada +5
Automatic Speech Recognition (ASR) is widely used within consumer devices such as mobile phones. Recently, personalization or on-device model fine-tuning has shown that adaptation…
persoDA: Personalized Data Augmentation for Personalized ASR
Pablo Peso Parada, Spyros Fontalis, Md Asif Jalal +6
Data augmentation (DA) is ubiquitously used in training of Automatic Speech Recognition (ASR) models. DA offers increased data variability, robustness and generalization against di…
On-Device Speaker Anonymization of Acoustic Embeddings for ASR based onFlexible Location Gradient Reversal Layer
Md Asif Jalal, Pablo Peso Parada, Jisi Zhang +5
Smart devices serviced by large-scale AI models necessitates user data transfer to the cloud for inference. For speech applications, this means transferring private user informatio…
FedNST: Federated Noisy Student Training for Automatic Speech Recognition
Haaris Mehmood, Agnieszka Dobrowolska, Karthikeyan Saravanan +1
Federated Learning (FL) enables training state-of-the-art Automatic Speech Recognition (ASR) models on user devices (clients) in distributed systems, hence preventing transmission…
pMCT: Patched Multi-Condition Training for Robust Speech Recognition
Pablo Peso Parada, Agnieszka Dobrowolska, Karthikeyan Saravanan +1
We propose a novel Patched Multi-Condition Training (pMCT) method for robust Automatic Speech Recognition (ASR). pMCT employs Multi-condition Audio Modification and Patching (MAMP)…