Ultra2Speech - A Deep Learning Framework for Formant Frequency Estimation and Tracking from Ultrasound Tongue Images

Saha, Pramit; Liu, Yadong; Gick, Bryan; Fels, Sidney

doi:10.1007/978-3-030-59716-0_45

Cited by 10 publications

(11 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Formally, our networks has to map each MRI image to a spectral vector. However, using several consecutive input frames instead of a single frame can significantly improve the results [6], [32]. Hence, the input for all our network configurations was a 3D array, treating time as the the third axis besides the two spacial axes of the images.…”

Section: D-cnn+bilstmmentioning

confidence: 99%

“…3D-CNN: Several authors argued recently that good video classification is also achievable using purely convolutional structures by extending the convolution to the temporal axis [6], [32], [36]. This is why a 3D convolutional model served as our second model (see Table I).…”

Section: D-cnn+bilstmmentioning

confidence: 99%

“…During the last decade, several types of devices have been proposed to record the movement of the articulators. The most important of these are ultrasound tongue imaging (UTI) [1]- [6], electromagnetic articulography (EMA) [7]- [10], permanent magnetic articulography (PMA) [11], [12], and surface electromyography (sEMG) [13]- [17]. While articulatory-toacoustic mapping can also be considered as a theoretical problem, it has an important application, which is the creation of silent speech interfaces [18].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Reconstructing Speech from Real-Time Articulatory MRI Using Neural Vocoders

Yu¹,

Shandiz²,

Tóth³

2021

Preprint

View full text Add to dashboard Cite

Several approaches exist for the recording of articulatory movements, such as eletromagnetic and permanent magnetic articulagraphy, ultrasound tongue imaging and surface electromyography. Although magnetic resonance imaging (MRI) is more costly than the above approaches, the recent developments in this area now allow the recording of real-time MRI videos of the articulators with an acceptable resolution.Here, we experiment with the reconstruction of the speech signal from a real-time MRI recording using deep neural networks. Instead of estimating speech directly, our networks are trained to output a spectral vector, from which we reconstruct the speech signal using the WaveGlow neural vocoder. We compare the performance of three deep neural architectures for the estimation task, combining convolutional (CNN) and recurrence-based (LSTM) neural layers. Besides the mean absolute error (MAE) of our networks, we also evaluate our models by comparing the speech signals obtained using several objective speech quality metrics like the mean cepstral distortion (MCD), Short-Time Objective Intelligibility (STOI), Perceptual Evaluation of Speech Quality (PESQ) and Signal-to-Distortion Ratio (SDR). The results indicate that our approach can successfully reconstruct the gross spectral shape, but more improvements are needed to reproduce the fine spectral details.

show abstract

Section: D-cnn+bilstmmentioning

confidence: 99%

Section: D-cnn+bilstmmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Reconstructing Speech from Real-Time Articulatory MRI Using Neural Vocoders

Yu¹,

Shandiz²,

Tóth³

2021

Preprint

View full text Add to dashboard Cite

show abstract

“…While this simple arrangement already performs reasonably well, significant improvement can be achieved by involving the input context, that is, by using a block of video frames as input instead of just one image. Several network architectures have been proposed to process 3D blocks of input data, for video processing in general [24,25,26], and for ultrasound input in particular [12,14,27,16,28]. In the experimental section we will experiment both with 2D and 3D Convolutional Neural Networks (CNNs) for the mapping task.…”

Section: The Uti-to-speech Frameworkmentioning

confidence: 99%

Legal education at the crossroads of social responsibility and the development of individual competencies – experience at the University of Szeged (Hungary)

Tóth

Kálmán

2021

Acta Iuris Stetinensis

View full text Add to dashboard Cite

The authors describe the role of the legal clinical education as an instrument for community empowerment through pro bono legal counseling, easily understandable e-compilation (Vademecum) of legal terms, and extended legal practice for law students, together with the development of their professional competencies. The new requirements in legal education, as determined by the government in 2016, focus on labour market needs, but academics and leaders of the University of Szeged have created an amalgam of tools for access to justice for local residents and NGOs in a less wealthy social environment, thereby introducing changes 1 The research for this study was carried out with the support of the programs of the Hungarian Ministry of Justice enhancing the standards of legal education.

show abstract

“…The goal is to convert this recording of the articulatory movement into a speech signal. Many possible approaches exist for this, but the most recent studies all apply deep neural networks (DNNs) for this task [23,4,21,7,27], and here we also apply neural structures that combine convolutional neural network (CNN) layers and recurrent layers such as the long short-term memory (LSTM) layer.…”

Section: Introductionmentioning

confidence: 99%

Voice Activity Detection for Ultrasound-Based Silent Speech Interfaces Using Convolutional Neural Networks

Shandiz¹,

Tóth²

2021

Text, Speech, and Dialogue

View full text Add to dashboard Cite

Voice Activity Detection (VAD) is not easy task when the input audio signal is noisy, and it is even more complicated when the input is not even an audio recording. This is the case with Silent Speech Interfaces (SSI) where we record the movement of the articulatory organs during speech, and we aim to reconstruct the speech signal from this recording. Our SSI system synthesizes speech from ultrasonic videos of the tongue movement, and the quality of the resulting speech signals are evaluated by metrics such as the mean squared error loss function of the underlying neural network and the Mel-Cepstral Distortion (MCD) of the reconstructed speech compared to the original. Here, we first demonstrate that the amount of silence in the training data can have an influence both on the MCD evaluation metric and on the performance of the neural network model. Then, we train a convolutional neural network classifier to separate silent and speech-containing ultrasound tongue images, using a conventional VAD algorithm to create the training labels from the corresponding speech signal. In the experiments our ultrasound-based speech/silence separator achieved a classification accuracy of about 85% and an AUC score around 86%.

show abstract

Ultra2Speech - A Deep Learning Framework for Formant Frequency Estimation and Tracking from Ultrasound Tongue Images

Cited by 10 publications

References 19 publications

Reconstructing Speech from Real-Time Articulatory MRI Using Neural Vocoders

Reconstructing Speech from Real-Time Articulatory MRI Using Neural Vocoders

Legal education at the crossroads of social responsibility and the development of individual competencies – experience at the University of Szeged (Hungary)

Voice Activity Detection for Ultrasound-Based Silent Speech Interfaces Using Convolutional Neural Networks

Contact Info

Product

Resources

About