Jean-Marc Valin
Research Subjects
Jean-Marc Valin has a total of 89 co-authors. Here are the top co-authors by number of shared publications:
F. Michaud
12 shared publications- Hiroshi G. Okuno
Future Robotics Organization, Waseda University, Okubo, Shinjuku-ku, Tokyo 169-0072, Japan
10 shared publications D. Letourneau
Dept. of Electr. Eng. & Comput. Eng., Sherbrooke Univ., Que., Canada
7 shared publicationsJ. Rouat
7 shared publications- Tetsuya Ogata
Department of Intermedia Art and Science, Waseda University, Tokyo, Japan
7 shared publications Francois Michaud
Department of Electrical Engineering and Computer Engineering Université de Sherbrooke 2500 boul
5 shared publications
Latest Publications
- 17 October 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Selected Topics in Signal Processing
- Vol. 18 (8) , 1441-1447
DRED: Deep REDundancy Coding of Speech Using a Rate-Distortion-Optimized Variational Autoencoder
Despite recent advancements in packet loss concealment (PLC) using deep learning techniques, packet loss remains a significant challenge in real-time speech communication. Redundancy has been used in the past to recover the missing information during losses. However, conventional redundancy techniques are limited in the maximum loss duration they can cover and are often unsuitable for burst packet loss. We propose a new approach based on a rate-distortion-optimized variational autoencoder (RDO-VAE), allowing us to optimize a deep speech compression algorithm for the task of encoding large amounts of redundancy at very low bitrate. The proposed Deep REDundancy (DRED) algorithm can transmit up to 50x redundancy using less than 32 kb/s. Results show that DRED outperforms the existing Opus codec redundancy. We also demonstrate its benefits when operating in the context of WebRTC.
- 17 October 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Selected Topics in Signal Processing
- Vol. 18 (8) , 1441-1447
DRED: Deep REDundancy Coding of Speech Using a Rate-Distortion-Optimized Variational Autoencoder
Despite recent advancements in packet loss concealment (PLC) using deep learning techniques, packet loss remains a significant challenge in real-time speech communication. Redundancy has been used in the past to recover the missing information during losses. However, conventional redundancy techniques are limited in the maximum loss duration they can cover and are often unsuitable for burst packet loss. We propose a new approach based on a rate-distortion-optimized variational autoencoder (RDO-VAE), allowing us to optimize a deep speech compression algorithm for the task of encoding large amounts of redundancy at very low bitrate. The proposed Deep REDundancy (DRED) algorithm can transmit up to 50x redundancy using less than 32 kb/s. Results show that DRED outperforms the existing Opus codec redundancy. We also demonstrate its benefits when operating in the context of WebRTC.
- 9 August 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Signal Processing Letters
- Vol. 31 (10709908) , 2115-2119
Very Low Complexity Speech Synthesis Using Framewise Autoregressive GAN (FARGAN) With Pitch Prediction
Neural vocoders are now being used in a wide range of speech processing applications. In many of those applications, thevocoder can be the mostcomplex component, so finding lower complexity algorithms can lead to significant practical benefits. In this work, we propose FARGAN, an autoregressive vocoder that takes advantage of long-term pitch prediction to synthesize high-quality speech in small subframes, without the need for teacher-forcing. Experimental results show that the proposed 600 MFLOPS FARGAN vocoder can achieve both higher quality and lower complexity than existing low-complexity vocoders. The quality even matches that of existing higher-complexity vocoders
- 9 August 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Signal Processing Letters
- Vol. 31 (10709908) , 2115-2119
Very Low Complexity Speech Synthesis Using Framewise Autoregressive GAN (FARGAN) With Pitch Prediction
Neural vocoders are now being used in a wide range of speech processing applications. In many of those applications, thevocoder can be the mostcomplex component, so finding lower complexity algorithms can lead to significant practical benefits. In this work, we propose FARGAN, an autoregressive vocoder that takes advantage of long-term pitch prediction to synthesize high-quality speech in small subframes, without the need for teacher-forcing. Experimental results show that the proposed 600 MFLOPS FARGAN vocoder can achieve both higher quality and lower complexity than existing low-complexity vocoders. The quality even matches that of existing higher-complexity vocoders
- Open Access
Noise-Robust DSP-Assisted Neural Pitch Estimation with Very Low Complexity
Pitch estimation is an essential step of many speech processing algorithms, including speech coding, synthesis, and enhancement. Recently, pitch estimators based on deep neural networks (DNNs) have have been outperforming well-established DSP-based techniques. Unfortunately, these new estimators can be impractical to deploy in real-time systems, both because of their relatively high complexity, and the fact that some require significant lookahead. We show that a hybrid estimator using a small deep neural network (DNN) with traditional DSP-based features can match or exceed the performance of pure DNN-based models, with a complexity and algorithmic delay comparable to traditional DSP-based algorithms. We further demonstrate that this hybrid approach can provide benefits for a neural vocoding task.
Open AccessNoise-Robust DSP-Assisted Neural Pitch Estimation with Very Low Complexity
Pitch estimation is an essential step of many speech processing algorithms, including speech coding, synthesis, and enhancement. Recently, pitch estimators based on deep neural networks (DNNs) have have been outperforming well-established DSP-based techniques. Unfortunately, these new estimators can be impractical to deploy in real-time systems, both because of their relatively high complexity, and the fact that some require significant lookahead. We show that a hybrid estimator using a small deep neural network (DNN) with traditional DSP-based features can match or exceed the performance of pure DNN-based models, with a complexity and algorithmic delay comparable to traditional DSP-based algorithms. We further demonstrate that this hybrid approach can provide benefits for a neural vocoding task.
- Open Access
NoLACE: Improving Low-Complexity Speech Codec Enhancement Through Adaptive Temporal Shaping
Speech codec enhancement methods are designed to remove distortions added by speech codecs. While classical methods are very low in complexity and add zero delay, their effectiveness is rather limited. Compared to that, DNN-based methods deliver higher quality but they are typically high in complexity and/or require delay. The recently proposed Linear Adaptive Coding Enhancer (LACE) addresses this problem by combining DNNs with classical long-term/short-term postfiltering resulting in a causal low-complexity model. A short-coming of the LACE model is, however, that quality quickly saturates when the model size is scaled up. To mitigate this problem, we propose a novel adatpive temporal shaping module that adds high temporal resolution to the LACE model resulting in the Non-Linear Adaptive Coding Enhancer (NoLACE). We adapt NoLACE to enhance the Opus codec and show that NoLACE significantly outperforms both the Opus baseline and an enlarged LACE model at 6, 9 and 12 kb/s. We also show that LACE and NoLACE are well-behaved when used with an ASR system.
Open AccessNoLACE: Improving Low-Complexity Speech Codec Enhancement Through Adaptive Temporal Shaping
Speech codec enhancement methods are designed to remove distortions added by speech codecs. While classical methods are very low in complexity and add zero delay, their effectiveness is rather limited. Compared to that, DNN-based methods deliver higher quality but they are typically high in complexity and/or require delay. The recently proposed Linear Adaptive Coding Enhancer (LACE) addresses this problem by combining DNNs with classical long-term/short-term postfiltering resulting in a causal low-complexity model. A short-coming of the LACE model is, however, that quality quickly saturates when the model size is scaled up. To mitigate this problem, we propose a novel adatpive temporal shaping module that adds high temporal resolution to the LACE model resulting in the Non-Linear Adaptive Coding Enhancer (NoLACE). We adapt NoLACE to enhance the Opus codec and show that NoLACE significantly outperforms both the Opus baseline and an enlarged LACE model at 6, 9 and 12 kb/s. We also show that LACE and NoLACE are well-behaved when used with an ASR system.
- Open Access
LACE: A light-weight, causal model for enhancing coded speech through adaptive convolutions
Classical speech coding uses low-complexity postfilters with zero lookahead to enhance the quality of coded speech, but their effectiveness is limited by their simplicity. Deep Neural Networks (DNNs) can be much more effective, but require high complexity and model size, or added delay. We propose a DNN model that generates classical filter kernels on a per-frame basis with a model of just 300~K parameters and 100~MFLOPS complexity, which is a practical complexity for desktop or mobile device CPUs. The lack of added delay allows it to be integrated into the Opus codec, and we demonstrate that it enables effective wideband encoding for bitrates down to 6 kb/s.
Open AccessLACE: A light-weight, causal model for enhancing coded speech through adaptive convolutions
Classical speech coding uses low-complexity postfilters with zero lookahead to enhance the quality of coded speech, but their effectiveness is limited by their simplicity. Deep Neural Networks (DNNs) can be much more effective, but require high complexity and model size, or added delay. We propose a DNN model that generates classical filter kernels on a per-frame basis with a model of just 300~K parameters and 100~MFLOPS complexity, which is a practical complexity for desktop or mobile device CPUs. The lack of added delay allows it to be integrated into the Opus codec, and we demonstrate that it enables effective wideband encoding for bitrates down to 6 kb/s.
- Open Access
Framewise WaveGAN: High Speed Adversarial Vocoder in Time Domain with Very Low Computational Complexity
GAN vocoders are currently one of the state-of-the-art methods for building high-quality neural waveform generative models. However, most of their architectures require dozens of billion floating-point operations per second (GFLOPS) to generate speech waveforms in samplewise manner. This makes GAN vocoders still challenging to run on normal CPUs without accelerators or parallel computers. In this work, we propose a new architecture for GAN vocoders that mainly depends on recurrent and fully-connected networks to directly generate the time domain signal in framewise manner. This results in considerable reduction of the computational cost and enables very fast generation on both GPUs and low-complexity CPUs. Experimental results show that our Framewise WaveGAN vocoder achieves significantly higher quality than auto-regressive maximum-likelihood vocoders such as LPCNet at a very low complexity of 1.2 GFLOPS. This makes GAN vocoders more practical on edge and low-power devices.
Open AccessFramewise WaveGAN: High Speed Adversarial Vocoder in Time Domain with Very Low Computational Complexity
GAN vocoders are currently one of the state-of-the-art methods for building high-quality neural waveform generative models. However, most of their architectures require dozens of billion floating-point operations per second (GFLOPS) to generate speech waveforms in samplewise manner. This makes GAN vocoders still challenging to run on normal CPUs without accelerators or parallel computers. In this work, we propose a new architecture for GAN vocoders that mainly depends on recurrent and fully-connected networks to directly generate the time domain signal in framewise manner. This results in considerable reduction of the computational cost and enables very fast generation on both GPUs and low-complexity CPUs. Experimental results show that our Framewise WaveGAN vocoder achieves significantly higher quality than auto-regressive maximum-likelihood vocoders such as LPCNet at a very low complexity of 1.2 GFLOPS. This makes GAN vocoders more practical on edge and low-power devices.
- Open Access
Low-Bitrate Redundancy Coding of Speech Using a Rate-Distortion-Optimized Variational Autoencoder
Robustness to packet loss is one of the main ongoing challenges in real-time speech communication. Deep packet loss concealment (PLC) techniques have recently demonstrated improved quality compared to traditional PLC. Despite that, all PLC techniques hit fundamental limitations when too much acoustic information is lost. To reduce losses in the first place, data is commonly sent multiple times using various redundancy mechanisms. We propose a neural speech coder specifically optimized to transmit a large amount of overlapping redundancy at a very low bitrate, up to 50x redundancy using less than 32~kb/s. Results show that the proposed redundancy is more effective than the existing Opus codec redundancy, and that the two can be combined for even greater robustness.
Open AccessLow-Bitrate Redundancy Coding of Speech Using a Rate-Distortion-Optimized Variational Autoencoder
Robustness to packet loss is one of the main ongoing challenges in real-time speech communication. Deep packet loss concealment (PLC) techniques have recently demonstrated improved quality compared to traditional PLC. Despite that, all PLC techniques hit fundamental limitations when too much acoustic information is lost. To reduce losses in the first place, data is commonly sent multiple times using various redundancy mechanisms. We propose a neural speech coder specifically optimized to transmit a large amount of overlapping redundancy at a very low bitrate, up to 50x redundancy using less than 32~kb/s. Results show that the proposed redundancy is more effective than the existing Opus codec redundancy, and that the two can be combined for even greater robustness.
- Open Access
End-to-end LPCNet: A Neural Vocoder With Fully-Differentiable LPC Estimation
Neural vocoders have recently demonstrated high quality speech synthesis, but typically require a high computational complexity. LPCNet was proposed as a way to reduce the complexity of neural synthesis by using linear prediction (LP) to assist an autoregressive model. At inference time, LPCNet relies on the LP coefficients being explicitly computed from the input acoustic features. That makes the design of LPCNet-based systems more complicated, while adding the constraint that the input features must represent a clean speech spectrum. We propose an end-to-end version of LPCNet that lifts these limitations by learning to infer the LP coefficients from the input features in the frame rate network. Results show that the proposed end-to-end approach equals or exceeds the quality of the original LPCNet model, but without explicit LP analysis. Our open-source end-to-end model still benefits from LPCNet's low complexity, while allowing for any type of conditioning features.
Open AccessEnd-to-end LPCNet: A Neural Vocoder With Fully-Differentiable LPC Estimation
Neural vocoders have recently demonstrated high quality speech synthesis, but typically require a high computational complexity. LPCNet was proposed as a way to reduce the complexity of neural synthesis by using linear prediction (LP) to assist an autoregressive model. At inference time, LPCNet relies on the LP coefficients being explicitly computed from the input acoustic features. That makes the design of LPCNet-based systems more complicated, while adding the constraint that the input features must represent a clean speech spectrum. We propose an end-to-end version of LPCNet that lifts these limitations by learning to infer the LP coefficients from the input features in the frame rate network. Results show that the proposed end-to-end approach equals or exceeds the quality of the original LPCNet model, but without explicit LP analysis. Our open-source end-to-end model still benefits from LPCNet's low complexity, while allowing for any type of conditioning features.
- Open Access
Neural Speech Synthesis on a Shoestring: Improving the Efficiency of LPCNet
Neural speech synthesis models can synthesize high quality speech but typically require a high computational complexity to do so. In previous work, we introduced LPCNet, which uses linear prediction to significantly reduce the complexity of neural synthesis. In this work, we further improve the efficiency of LPCNet -- targeting both algorithmic and computational improvements -- to make it usable on a wide variety of devices. We demonstrate an improvement in synthesis quality while operating 2.5x faster. The resulting open-source LPCNet algorithm can perform real-time neural synthesis on most existing phones and is even usable in some embedded devices.
Open AccessNeural Speech Synthesis on a Shoestring: Improving the Efficiency of LPCNet
Neural speech synthesis models can synthesize high quality speech but typically require a high computational complexity to do so. In previous work, we introduced LPCNet, which uses linear prediction to significantly reduce the complexity of neural synthesis. In this work, we further improve the efficiency of LPCNet -- targeting both algorithmic and computational improvements -- to make it usable on a wide variety of devices. We demonstrate an improvement in synthesis quality while operating 2.5x faster. The resulting open-source LPCNet algorithm can perform real-time neural synthesis on most existing phones and is even usable in some embedded devices.
- Open Access
A Perceptually-Motivated Approach for Low-Complexity, Real-Time Enhancement of Fullband Speech
Over the past few years, speech enhancement methods based on deep learning have greatly surpassed traditional methods based on spectral subtraction and spectral estimation. Many of these new techniques operate directly in the the short-time Fourier transform (STFT) domain, resulting in a high computational complexity. In this work, we propose PercepNet, an efficient approach that relies on human perception of speech by focusing on the spectral envelope and on the periodicity of the speech. We demonstrate high-quality, real-time enhancement of fullband (48 kHz) speech with less than 5% of a CPU core.
Open AccessA Perceptually-Motivated Approach for Low-Complexity, Real-Time Enhancement of Fullband Speech
Over the past few years, speech enhancement methods based on deep learning have greatly surpassed traditional methods based on spectral subtraction and spectral estimation. Many of these new techniques operate directly in the the short-time Fourier transform (STFT) domain, resulting in a high computational complexity. In this work, we propose PercepNet, an efficient approach that relies on human perception of speech by focusing on the spectral envelope and on the periodicity of the speech. We demonstrate high-quality, real-time enhancement of fullband (48 kHz) speech with less than 5% of a CPU core.