Tamás Gábor Csapó

Milán András Fodor, Tamás Gábor Csapó, Frigyes Viktor Arthur

The aim of the study is to investigate the complex mechanisms of speech perception and ultimately decode the electrical changes in the brain accruing while listening to speech. We attempt to decode heard speech from intracranial electroencephalographic (iEEG) data using deep learning methods. The goal is to aid the advancement of brain-computer interface (BCI) technology for speech synthesis, and, hopefully, to provide an additional perspective on the cognitive processes of speech perception. This approach diverges from the conventional focus on speech production and instead chooses to investigate neural representations of perceived speech. This angle opened up a complex perspective, potentially allowing us to study more sophisticated neural patterns. Leveraging the power of deep learning models, the research aimed to establish a connection between these intricate neural activities and the corresponding speech sounds. Despite the approach not having achieved a breakthrough yet, the research sheds light on the potential of decoding neural activity during speech perception. Our current efforts can serve as a foundation, and we are optimistic about the potential of expanding and improving upon this work to move closer towards more advanced BCIs, better understanding of processes underlying perceived speech and its relation to spoken speech.

Ali Raheem Mandeel, Mohammed Salah Al-Radhi, Tamás Gábor Csapó

Traditional vocoder-based statistical parametric speech synthesis can be advantageous in applications that require low computational complexity. Recent neural vocoders, which can produce high naturalness, still cannot fulfill the requirement of being real-time during synthesis. In this paper, we experiment with our earlier continuous vocoder, in which the excitation is modeled with two one-dimensional parameters: continuous F0 and Maximum Voiced Frequency. We show on the data of 9 speakers that an average voice can be trained for DNN-TTS, and speaker adaptation is feasible 400 utterances (about 14 minutes). Objective experiments support that the quality of speaker adaptation with Continuous Vocoder-based DNN-TTS is similar to the quality of the speaker adaptation with a WORLD Vocoder-based baseline.

Csaba Zainkó, László Tóth, Amin Honarmandi Shandiz et al.

For articulatory-to-acoustic mapping, typically only limited parallel training data is available, making it impossible to apply fully end-to-end solutions like Tacotron2. In this paper, we experimented with transfer learning and adaptation of a Tacotron2 text-to-speech model to improve the final synthesis quality of ultrasound-based articulatory-to-acoustic mapping with a limited database. We use a multi-speaker pre-trained Tacotron2 TTS model and a pre-trained WaveGlow neural vocoder. The articulatory-to-acoustic conversion contains three steps: 1) from a sequence of ultrasound tongue image recordings, a 3D convolutional neural network predicts the inputs of the pre-trained Tacotron2 model, 2) the Tacotron2 model converts this intermediate representation to an 80-dimensional mel-spectrogram, and 3) the WaveGlow model is applied for final inference. This generated speech contains the timing of the original articulatory data from the ultrasound recording, but the F0 contour and the spectral information is predicted by the Tacotron2 model. The F0 values are independent of the original ultrasound images, but represent the target speaker, as they are inferred from the pre-trained Tacotron2 model. In our experiments, we demonstrated that the synthesized speech quality is more natural with the proposed solutions than with our earlier model.

Tamás Gábor Csapó

In this paper, we present our first experiments in text-to-articulation prediction, using ultrasound tongue image targets. We extend a traditional (vocoder-based) DNN-TTS framework with predicting PCA-compressed ultrasound images, of which the continuous tongue motion can be reconstructed in synchrony with synthesized speech. We use the data of eight speakers, train fully connected and recurrent neural networks, and show that FC-DNNs are more suitable for the prediction of sequential data than LSTMs, in case of limited training data. Objective experiments and visualized predictions show that the proposed solution is feasible and the generated ultrasound videos are close to natural tongue movement. Articulatory movement prediction from text input can be useful for audiovisual speech synthesis or computer-assisted pronunciation training.

Tamás Gábor Csapó, László Tóth, Gábor Gosztolya et al.

Articulatory information has been shown to be effective in improving the performance of HMM-based and DNN-based text-to-speech synthesis. Speech synthesis research focuses traditionally on text-to-speech conversion, when the input is text or an estimated linguistic representation, and the target is synthesized speech. However, a research field that has risen in the last decade is articulation-to-speech synthesis (with a target application of a Silent Speech Interface, SSI), when the goal is to synthesize speech from some representation of the movement of the articulatory organs. In this paper, we extend traditional (vocoder-based) DNN-TTS with articulatory input, estimated from ultrasound tongue images. We compare text-only, ultrasound-only, and combined inputs. Using data from eight speakers, we show that that the combined text and articulatory input can have advantages in limited-data scenarios, namely, it may increase the naturalness of synthesized speech compared to single text input. Besides, we analyze the ultrasound tongue recordings of several speakers, and show that misalignments in the ultrasound transducer positioning can have a negative effect on the final synthesis performance.

Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Géza Németh

Vocoders received renewed attention as main components in statistical parametric text-to-speech (TTS) synthesis and speech transformation systems. Even though there are vocoding techniques give almost accepted synthesized speech, their high computational complexity and irregular structures are still considered challenging concerns, which yield a variety of voice quality degradation. Therefore, this paper presents new techniques in a continuous vocoder, that is all features are continuous and presents a flexible speech synthesis system. First, a new continuous noise masking based on the phase distortion is proposed to eliminate the perceptual impact of the residual noise and letting an accurate reconstruction of noise characteristics. Second, we addressed the need of neural sequence to sequence modeling approach for the task of TTS based on recurrent networks. Bidirectional long short-term memory (LSTM) and gated recurrent unit (GRU) are studied and applied to model continuous parameters for more natural-sounding like a human. The evaluation results proved that the proposed model achieves the state-of-the-art performance of the speech synthesis compared with the other traditional methods.

Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Csaba Zainkó et al.

To date, various speech technology systems have adopted the vocoder approach, a method for synthesizing speech waveform that shows a major role in the performance of statistical parametric speech synthesis. WaveNet one of the best models that nearly resembles the human voice, has to generate a waveform in a time consuming sequential manner with an extremely complex structure of its neural networks.

Amin Honarmandi Shandiz, László Tóth, Gábor Gosztolya et al.

Articulatory-to-acoustic mapping seeks to reconstruct speech from a recording of the articulatory movements, for example, an ultrasound video. Just like speech signals, these recordings represent not only the linguistic content, but are also highly specific to the actual speaker. Hence, due to the lack of multi-speaker data sets, researchers have so far concentrated on speaker-dependent modeling. Here, we present multi-speaker experiments using the recently published TaL80 corpus. To model speaker characteristics, we adjusted the x-vector framework popular in speech processing to operate with ultrasound tongue videos. Next, we performed speaker recognition experiments using 50 speakers from the corpus. Then, we created speaker embedding vectors and evaluated them on the remaining speakers. Finally, we examined how the embedding vector influences the accuracy of our ultrasound-to-speech conversion network in a multi-speaker scenario. In the experiments we attained speaker recognition error rates below 3%, and we also found that the embedding vectors generalize nicely to unseen speakers. Our first attempt to apply them in a multi-speaker silent speech framework brought about a marginal reduction in the error rate of the spectral estimation step.

Frigyes Viktor Arthur, Tamás Gábor Csapó

Articulatory-to-acoustic (forward) mapping is a technique to predict speech using various articulatory acquisition techniques as input (e.g. ultrasound tongue imaging, MRI, lip video). The advantage of lip video is that it is easily available and affordable: most modern smartphones have a front camera. There are already a few solutions for lip-to-speech synthesis, but they mostly concentrate on offline training and inference. In this paper, we propose a system built from a backend for deep neural network training and inference and a fronted as a form of a mobile application. Our initial evaluation shows that the scenario is feasible: a top-5 classification accuracy of 74% is combined with feedback from the mobile application user, making sure that the speaking impaired might be able to communicate with this solution.

Amin Honarmandi Shandiz, László Tóth, Gábor Gosztolya et al.

Besides the well-known classification task, these days neural networks are frequently being applied to generate or transform data, such as images and audio signals. In such tasks, the conventional loss functions like the mean squared error (MSE) may not give satisfactory results. To improve the perceptual quality of the generated signals, one possibility is to increase their similarity to real signals, where the similarity is evaluated via a discriminator network. The combination of the generator and discriminator nets is called a Generative Adversarial Network (GAN). Here, we evaluate this adversarial training framework in the articulatory-to-acoustic mapping task, where the goal is to reconstruct the speech signal from a recording of the movement of articulatory organs. As the generator, we apply a 3D convolutional network that gave us good results in an earlier study. To turn it into a GAN, we extend the conventional MSE training loss with an adversarial loss component provided by a discriminator network. As for the evaluation, we report various objective speech quality metrics such as the Perceptual Evaluation of Speech Quality (PESQ), and the Mel-Cepstral Distortion (MCD). Our results indicate that the application of the adversarial training loss brings about a slight, but consistent improvement in all these metrics.

Tamás Gábor Csapó, Kele Xu

In speech production research, different imaging modalities have been employed to obtain accurate information about the movement and shaping of the vocal tract. Ultrasound is an affordable and non-invasive imaging modality with relatively high temporal and spatial resolution to study the dynamic behavior of tongue during speech production. However, a long-standing problem for ultrasound tongue imaging is the transducer misalignment during longer data recording sessions. In this paper, we propose a simple, yet effective, misalignment quantification approach. The analysis employs MSE distance and two similarity measurement metrics to identify the relative displacement between the chin and the transducer. We visualize these measures as a function of the timestamp of the utterances. Extensive experiments are conducted on a Hungarian and Scottish English child dataset. The results suggest that large values of Mean Square Error (MSE) and small values of Structural Similarity Index (SSIM) and Complex Wavelet SSIM indicate corruptions or issues during the data recordings, which can either be caused by transducer misalignment or lack of gel.

Tamás Gábor Csapó, Csaba Zainkó, László Tóth et al.

For articulatory-to-acoustic mapping using deep neural networks, typically spectral and excitation parameters of vocoders have been used as the training targets. However, vocoding often results in buzzy and muffled final speech quality. Therefore, in this paper on ultrasound-based articulatory-to-acoustic conversion, we use a flow-based neural vocoder (WaveGlow) pre-trained on a large amount of English and Hungarian speech data. The inputs of the convolutional neural network are ultrasound tongue images. The training target is the 80-dimensional mel-spectrogram, which results in a finer detailed spectral representation than the previously used 25-dimensional Mel-Generalized Cepstrum. From the output of the ultrasound-to-mel-spectrogram prediction, WaveGlow inference results in synthesized speech. We compare the proposed WaveGlow-based system with a continuous vocoder which does not use strict voiced/unvoiced decision when predicting F0. The results demonstrate that during the articulatory-to-acoustic mapping experiments, the WaveGlow neural vocoder produces significantly more natural synthesized speech than the baseline system. Besides, the advantage of WaveGlow is that F0 is included in the mel-spectrogram representation, and it is not necessary to predict the excitation separately.

Tamás Gábor Csapó

Acoustic-to-articulatory inversion (AAI) methods estimate articulatory movements from the acoustic speech signal, which can be useful in several tasks such as speech recognition, synthesis, talking heads and language tutoring. Most earlier inversion studies are based on point-tracking articulatory techniques (e.g. EMA or XRMB). The advantage of rtMRI is that it provides dynamic information about the full midsagittal plane of the upper airway, with a high 'relative' spatial resolution. In this work, we estimated midsagittal rtMRI images of the vocal tract for speaker dependent AAI, using MGC-LSP spectral features as input. We applied FC-DNNs, CNNs and recurrent neural networks, and have shown that LSTMs are the most suitable for this task. As objective evaluation we measured normalized MSE, Structural Similarity Index (SSIM) and its complex wavelet version (CW-SSIM). The results indicate that the combination of FC-DNNs and LSTMs can achieve smooth generated MR images of the vocal tract, which are similar to the original MRI recordings (average CW-SSIM: 0.94).

Tamás Gábor Csapó

Articulatory-to-acoustic (forward) mapping is a technique to predict speech using various articulatory acquisition techniques (e.g. ultrasound tongue imaging, lip video). Real-time MRI (rtMRI) of the vocal tract has not been used before for this purpose. The advantage of MRI is that it has a high `relative' spatial resolution: it can capture not only lingual, labial and jaw motion, but also the velum and the pharyngeal region, which is typically not possible with other techniques. In the current paper, we train various DNNs (fully connected, convolutional and recurrent neural networks) for articulatory-to-speech conversion, using rtMRI as input, in a speaker-specific way. We use two male and two female speakers of the USC-TIMIT articulatory database, each of them uttering 460 sentences. We evaluate the results with objective (Normalized MSE and MCD) and subjective measures (perceptual test) and show that CNN-LSTM networks are preferred which take multiple images as input, and achieve MCD scores between 2.8-4.5 dB. In the experiments, we find that the predictions of speaker `m1' are significantly weaker than other speakers. We show that this is caused by the fact that 74% of the recordings of speaker `m1' are out of sync.

Tamás Gábor Csapó, Mohammed Salah Al-Radhi, Géza Németh et al.

Recently it was shown that within the Silent Speech Interface (SSI) field, the prediction of F0 is possible from Ultrasound Tongue Images (UTI) as the articulatory input, using Deep Neural Networks for articulatory-to-acoustic mapping. Moreover, text-to-speech synthesizers were shown to produce higher quality speech when using a continuous pitch estimate, which takes non-zero pitch values even when voicing is not present. Therefore, in this paper on UTI-based SSI, we use a simple continuous F0 tracker which does not apply a strict voiced / unvoiced decision. Continuous vocoder parameters (ContF0, Maximum Voiced Frequency and Mel-Generalized Cepstrum) are predicted using a convolutional neural network, with UTI as input. The results demonstrate that during the articulatory-to-acoustic mapping experiments, the continuous F0 is predicted with lower error, and the continuous vocoder produces slightly more natural synthesized speech than the baseline vocoder using standard discontinuous F0.

Dagoberto Porras, Alexander Sepúlveda-Sepúlveda, Tamás Gábor Csapó

Speech sounds are produced as the coordinated movement of the speaking organs. There are several available methods to model the relation of articulatory movements and the resulting speech signal. The reverse problem is often called as acoustic-to-articulatory inversion (AAI). In this paper we have implemented several different Deep Neural Networks (DNNs) to estimate the articulatory information from the acoustic signal. There are several previous works related to performing this task, but most of them are using ElectroMagnetic Articulography (EMA) for tracking the articulatory movement. Compared to EMA, Ultrasound Tongue Imaging (UTI) is a technique of higher cost-benefit if we take into account equipment cost, portability, safety and visualized structures. Seeing that, our goal is to train a DNN to obtain UT images, when using speech as input. We also test two approaches to represent the articulatory information: 1) the EigenTongue space and 2) the raw ultrasound image. As an objective quality measure for the reconstructed UT images, we use MSE, Structural Similarity Index (SSIM) and Complex-Wavelet SSIM (CW-SSIM). Our experimental results show that CW-SSIM is the most useful error measure in the UTI context. We tested three different system configurations: a) simple DNN composed of 2 hidden layers with 64x64 pixels of an UTI file as target; b) the same simple DNN but with ultrasound images projected to the EigenTongue space as the target; c) and a more complex DNN composed of 5 hidden layers with UTI files projected to the EigenTongue space. In a subjective experiment the subjects found that the neural networks with two hidden layers were more suitable for this inversion task.

Mohammed Salah Al-Radhi, Tamás Gábor Csapó, Géza Németh

Recently in statistical parametric speech synthesis, we proposed a continuous sinusoidal model (CSM) using continuous F0 (contF0) in combination with Maximum Voiced Frequency (MVF), which was successfully giving state-of-the-art vocoders performance (e.g. similar to STRAIGHT) in synthesized speech. In this paper, we address the use of sequence-to-sequence modeling with recurrent neural networks (RNNs). Bidirectional long short-term memory (Bi-LSTM) is investigated and applied using our CSM to model contF0, MVF, and Mel-Generalized Cepstrum (MGC) for more natural sounding synthesized speech. For refining the output of the contF0 estimation, post-processing based on time-warping approach is applied to reduce the unwanted voiced component of the unvoiced speech sounds, resulting in an enhanced contF0 track. The overall conclusion is covered by objective evaluation and subjective listening test, showing that the proposed framework provides satisfactory results in terms of naturalness and intelligibility, and is comparable to the high-quality WORLD model based RNNs.

Gábor Gosztolya, Ádám Pintér, László Tóth et al.

When using ultrasound video as input, Deep Neural Network-based Silent Speech Interfaces usually rely on the whole image to estimate the spectral parameters required for the speech synthesis step. Although this approach is quite straightforward, and it permits the synthesis of understandable speech, it has several disadvantages as well. Besides the inability to capture the relations between close regions (i.e. pixels) of the image, this pixel-by-pixel representation of the image is also quite uneconomical. It is easy to see that a significant part of the image is irrelevant for the spectral parameter estimation task as the information stored by the neighbouring pixels is redundant, and the neural network is quite large due to the large number of input features. To resolve these issues, in this study we train an autoencoder neural network on the ultrasound image; the estimation of the spectral speech parameters is done by a second DNN, using the activations of the bottleneck layer of the autoencoder network as features. In our experiments, the proposed method proved to be more efficient than the standard approach: the measured normalized mean squared error scores were lower, while the correlation values were higher in each case. Based on the result of a listening test, the synthesized utterances also sounded more natural to native speakers. A further advantage of our proposed approach is that, due to the (relatively) small size of the bottleneck layer, we can utilize several consecutive ultrasound images during estimation without a significant increase in the network size, while significantly increasing the accuracy of parameter estimation.