arXiv subjects

Lyonel Behringer

Publications and source records attributed to Lyonel Behringer.

4 recordsLinked to original sources

arXiv2026

Assessing the Impact of Noise and Speech Enhancement on the Intelligibility of Speech Codecs

Preserving speech intelligibility is a minimum requirement for speech codecs in communication. Recently, very low-bitrate neural codecs have gained interest for replacing classical codecs, reinforcing the need to evaluate whether intelligibility is preserved in realistic scenarios. In this paper, we evaluate the intelligibility and listening effort of classical and neural speech codecs in clean and noisy conditions. Further, we assess the impact of speech enhancement (SE) before coding, simulating a possible audio processing pipeline. The results show that classical codecs are more noise robust than neural codecs. Further, SE can lead to significant intelligibility and listening effort improvements for codecs otherwise negatively affected by noise. Listening effort reveals nuanced differences when intelligibility is saturated. Lastly, objective intelligibility based on automatic speech recognition is highly correlated with subjective intelligibility scores averaged per condition.

eess.AS↗

arXiv2025

Benchmarking Neural Speech Codec Intelligibility with SITool

Speech intelligibility assessment is essential for evaluating neural speech codecs, yet most evaluation efforts focus on overall quality rather than intelligibility. Only a few publicly available tools exist for conducting standardized intelligibility tests, like the Diagnostic Rhyme Test (DRT) and Modified Rhyme Test (MRT). We introduce the Speech Intelligibility Toolkit for Subjective Evaluation (SITool), a Flask-based web application for conducting DRT and MRT in laboratory and crowdsourcing settings. We use SITool to benchmark 13 neural and traditional speech codecs, analyzing phoneme-level degradations and comparing subjective DRT results with objective intelligibility metrics. Our findings show that, while neural speech codecs can outperform traditional ones in subjective intelligibility, only STOI and ESTOI - not WER - significantly correlate with subjective results, although they struggle to capture gender and wordlist-specific variations observed in subjective evaluations.

eess.AS↗

arXiv2024

Meta Learning Text-to-Speech Synthesis in over 7000 Languages

In this work, we take on the challenging task of building a single text-to-speech synthesis system that is capable of generating speech in over 7000 languages, many of which lack sufficient data for traditional TTS development. By leveraging a novel integration of massively multilingual pretraining and meta learning to approximate language representations, our approach enables zero-shot speech synthesis in languages without any available data. We validate our system's performance through objective measures and human evaluation across a diverse linguistic landscape. By releasing our code and models publicly, we aim to empower communities with limited linguistic resources and foster further innovation in the field of speech technology.

cs.CL↗

arXiv2024

Neural Speech Coding for Real-time Communications using Constant Bitrate Scalar Quantization

Neural audio coding has emerged as a vivid research direction by promising good audio quality at very low bitrates unachievable by classical coding techniques. Here, end-to-end trainable autoencoder-like models represent the state of the art, where a discrete representation in the bottleneck of the autoencoder is learned. This allows for efficient transmission of the input audio signal. The learned discrete representation of neural codecs is typically generated by applying a quantizer to the output of the neural encoder. In almost all state-of-the-art neural audio coding approaches, this quantizer is realized as a Vector Quantizer (VQ) and a lot of effort has been spent to alleviate drawbacks of this quantization technique when used together with a neural audio coder. In this paper, we propose and analyze simple alternatives to VQ, which are based on projected Scalar Quantization (SQ). These quantization techniques do not need any additional losses, scheduling parameters or codebook storage thereby simplifying the training of neural audio codecs. For real-time speech communication applications, these neural codecs are required to operate at low complexity, low latency and at low bitrates. We address those challenges by proposing a new causal network architecture that is based on SQ and a Short-Time Fourier Transform (STFT) representation. The proposed method performs particularly well in the very low complexity and low bitrate regime.

eess.AS↗