Picture for Hynek Hermansky

Hynek Hermansky

Self-supervised Learning with Speech Modulation Dropout

Mar 22, 2023
Samik Sadhu, Hynek Hermansky

Figure 1 for Self-supervised Learning with Speech Modulation Dropout

Figure 2 for Self-supervised Learning with Speech Modulation Dropout

Figure 3 for Self-supervised Learning with Speech Modulation Dropout

Figure 4 for Self-supervised Learning with Speech Modulation Dropout

Access Paper or Ask Questions

Blind Signal Dereverberation for Machine Speech Recognition

Sep 30, 2022
Samik Sadhu, Hynek Hermansky

Figure 1 for Blind Signal Dereverberation for Machine Speech Recognition

Figure 2 for Blind Signal Dereverberation for Machine Speech Recognition

Figure 3 for Blind Signal Dereverberation for Machine Speech Recognition

Figure 4 for Blind Signal Dereverberation for Machine Speech Recognition

Access Paper or Ask Questions

Importance of Different Temporal Modulations of Speech: A Tale of Two Perspectives

Mar 31, 2022
Samik Sadhu, Hynek Hermansky

Figure 1 for Importance of Different Temporal Modulations of Speech: A Tale of Two Perspectives

Figure 2 for Importance of Different Temporal Modulations of Speech: A Tale of Two Perspectives

Figure 3 for Importance of Different Temporal Modulations of Speech: A Tale of Two Perspectives

Figure 4 for Importance of Different Temporal Modulations of Speech: A Tale of Two Perspectives

* Submitted to INTERSPEECH 2022

Access Paper or Ask Questions

Complex Frequency Domain Linear Prediction: A Tool to Compute Modulation Spectrum of Speech

Mar 31, 2022
Samik Sadhu, Hynek Hermansky

Figure 1 for Complex Frequency Domain Linear Prediction: A Tool to Compute Modulation Spectrum of Speech

Figure 2 for Complex Frequency Domain Linear Prediction: A Tool to Compute Modulation Spectrum of Speech

Figure 3 for Complex Frequency Domain Linear Prediction: A Tool to Compute Modulation Spectrum of Speech

Figure 4 for Complex Frequency Domain Linear Prediction: A Tool to Compute Modulation Spectrum of Speech

* Submitted to INTERSPEECH 2022

Access Paper or Ask Questions

Radically Old Way of Computing Spectra: Applications in End-to-End ASR

Apr 02, 2021
Samik Sadhu, Hynek Hermansky

Figure 1 for Radically Old Way of Computing Spectra: Applications in End-to-End ASR

Figure 2 for Radically Old Way of Computing Spectra: Applications in End-to-End ASR

Figure 3 for Radically Old Way of Computing Spectra: Applications in End-to-End ASR

Figure 4 for Radically Old Way of Computing Spectra: Applications in End-to-End ASR

* submitted to INTERSPEECH 2021

Access Paper or Ask Questions

FDLP-Spectrogram: Capturing Speech Dynamics in Spectrograms for End-to-end Automatic Speech Recognition

Mar 25, 2021
Samik Sadhu, Hynek Hermansky

Figure 1 for FDLP-Spectrogram: Capturing Speech Dynamics in Spectrograms for End-to-end Automatic Speech Recognition

Figure 2 for FDLP-Spectrogram: Capturing Speech Dynamics in Spectrograms for End-to-end Automatic Speech Recognition

Figure 3 for FDLP-Spectrogram: Capturing Speech Dynamics in Spectrograms for End-to-end Automatic Speech Recognition

Figure 4 for FDLP-Spectrogram: Capturing Speech Dynamics in Spectrograms for End-to-end Automatic Speech Recognition

* submitted to INTERSPEECH

Access Paper or Ask Questions

Two-Stage Augmentation and Adaptive CTC Fusion for Improved Robustness of Multi-Stream End-to-End ASR

Feb 05, 2021
Ruizhi Li, Gregory Sell, Hynek Hermansky

Figure 1 for Two-Stage Augmentation and Adaptive CTC Fusion for Improved Robustness of Multi-Stream End-to-End ASR

Figure 2 for Two-Stage Augmentation and Adaptive CTC Fusion for Improved Robustness of Multi-Stream End-to-End ASR

Figure 3 for Two-Stage Augmentation and Adaptive CTC Fusion for Improved Robustness of Multi-Stream End-to-End ASR

Figure 4 for Two-Stage Augmentation and Adaptive CTC Fusion for Improved Robustness of Multi-Stream End-to-End ASR

* Accepted at IEEE SLT 2021

Access Paper or Ask Questions

A practical two-stage training strategy for multi-stream end-to-end speech recognition

Oct 23, 2019
Ruizhi Li, Gregory Sell, Xiaofei Wang, Shinji Watanabe, Hynek Hermansky

Figure 1 for A practical two-stage training strategy for multi-stream end-to-end speech recognition

Figure 2 for A practical two-stage training strategy for multi-stream end-to-end speech recognition

Figure 3 for A practical two-stage training strategy for multi-stream end-to-end speech recognition

Figure 4 for A practical two-stage training strategy for multi-stream end-to-end speech recognition

* submitted to ICASSP 2019

Access Paper or Ask Questions

Multi-Stream End-to-End Speech Recognition

Jun 17, 2019
Ruizhi Li, Xiaofei Wang, Sri Harish Mallidi, Shinji Watanabe, Takaaki Hori, Hynek Hermansky

Figure 1 for Multi-Stream End-to-End Speech Recognition

Figure 2 for Multi-Stream End-to-End Speech Recognition

Figure 3 for Multi-Stream End-to-End Speech Recognition

Figure 4 for Multi-Stream End-to-End Speech Recognition

* submitted to IEEE TASLP. arXiv admin note: substantial text overlap with arXiv:1811.04897, arXiv:1811.04903

Access Paper or Ask Questions