Improved Far-field Speech Recognition Using Joint Variational Autoencoder

Abstract

Automatic Speech Recognition (ASR) systems suffer considerably when source speech is corrupted with noise or room impulse responses (RIR). Typically, speech enhancement is applied in both mismatched and matched scenario training and testing. In matched setting, acoustic model (AM) is trained on dereverberated far-field features while in mismatched setting, AM is fixed. In recent past, mapping speech features from far-field to close-talk using denoising autoencoder (DA) has been explored. In this paper, we focus on matched scenario training and show that the proposed joint VAE based mapping achieves a significant improvement over DA. Specifically, we observe an absolute improvement of 2.5% in word error rate (WER) compared to DA based enhancement and 3.96% compared to AM trained directly on far-field filterbank features.

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keykumar2022improved

Related papers

• Single Channel Far Field Feature Enhancement For Speaker Verification In The Wild (2020)0.00
• Correlation Distance Skip Connection Denoising Autoencoder (CDSK-DAE) For Speech Feature Enhancement (2019)6.77
• Task-specific Optimization Of Virtual Channel Linear Prediction-based Speech Dereverberation Front-end For Far-field Speaker Verification (2021)2.26
• Frequency Domain Multi-channel Acoustic Modeling For Distant Speech Recognition (2019)9.92
• Dereverberation Of Autoregressive Envelopes For Far-field Speech Recognition (2021)6.77
• Leveraging Redundancy In Multiple Audio Signals For Far-field Speech Recognition (2023)0.00
• Investigation Of Speech And Noise Latent Representations In Single-channel Vae-based Speech Enhancement (2025)0.00
• Joint Beamforming And Speaker-attributed ASR For Real Distant-microphone Meeting Transcription (2024)2.26