On-Device Speaker Anonymization of Acoustic Embeddings for ASR based onFlexible Location Gradient Reversal Layer.
Md Asif JalalPablo Peso ParadaJisi ZhangKarthikeyan SaravananMete OzayMyoungji HanJungin LeeSeokyeong JungPublished in: CoRR (2023)
Keyphrases
- automatic speech recognition
- acoustic features
- speech recognition
- speech sounds
- speech recognizers
- acoustic models
- speech segments
- hidden markov models
- information loss
- speech signal
- speech recognition systems
- prosodic features
- noisy environments
- speaker verification
- multi layer
- speaker independent
- privacy preserving
- vector space
- broadcast news
- speaker identification
- euclidean space
- spontaneous speech
- vocal tract
- application layer
- speaker diarization
- location based services
- audio visual
- low dimensional
- pattern recognition