George Close

George Close, William Ravenscroft, Thomas Hain et al.

Recent work in the domain of speech enhancement has explored the use of self-supervised speech representations to aid in the training of neural speech enhancement models. However, much of this work focuses on using the deepest or final outputs of self supervised speech representation models, rather than the earlier feature encodings. The use of self supervised representations in such a way is often not fully motivated. In this work it is shown that the distance between the feature encodings of clean and noisy speech correlate strongly with psychoacoustically motivated measures of speech quality and intelligibility, as well as with human Mean Opinion Score (MOS) ratings. Experiments using this distance as a loss function are performed and improved performance over the use of STFT spectrogram distance based loss as well as other common loss functions from speech enhancement literature is demonstrated using objective measures such as perceptual evaluation of speech quality (PESQ) and short-time objective intelligibility (STOI).

George Close, Thomas Hain, Stefan Goetze

Training of speech enhancement systems often does not incorporate knowledge of human perception and thus can lead to unnatural sounding results. Incorporating psychoacoustically motivated speech perception metrics as part of model training via a predictor network has recently gained interest. However, the performance of such predictors is limited by the distribution of metric scores that appear in the training data. In this work, we propose MetricGAN+/- (an extension of MetricGAN+, one such metric-motivated system) which introduces an additional network - a "de-generator" which attempts to improve the robustness of the prediction network (and by extension of the generator) by ensuring observation of a wider range of metric scores in training. Experimental results on the VoiceBank-DEMAND dataset show relative improvement in PESQ score of 3.8% (3.05 vs 3.22 PESQ score), as well as better generalisation to unseen noise and speech.

George Close, Thomas Hain, Stefan Goetze

Recent work in the field of speech enhancement (SE) has involved the use of self-supervised speech representations (SSSRs) as feature transformations in loss functions. However, in prior work, very little attention has been paid to the relationship between the language of the audio used to train the self-supervised representation and that used to train the SE system. Enhancement models trained using a loss function which incorporates a self-supervised representation that shares exactly the language of the noisy data used to train the SE system show better performance than those which do not match exactly. This may lead to enhancement systems which are language specific and as such do not generalise well to unseen languages, unlike models trained using traditional spectrogram or time domain loss functions. In this work, SE models are trained and tested on a number of different languages, with self-supervised representations which themselves are trained using different language combinations and with differing network structures as loss function representations. These models are then tested across unseen languages and their performances are analysed. It is found that the training language of the self-supervised representation appears to have a minor effect on enhancement performance, the amount of training data of a particular language, however, greatly affects performance.

George Close, Thomas Hain, Stefan Goetze

Self-supervised speech representations (SSSRs) have been successfully applied to a number of speech-processing tasks, e.g. as feature extractor for speech quality (SQ) prediction, which is, in turn, relevant for assessment and training speech enhancement systems for users with normal or impaired hearing. However, exact knowledge of why and how quality-related information is encoded well in such representations remains poorly understood. In this work, techniques for non-intrusive prediction of SQ ratings are extended to the prediction of intelligibility for hearing-impaired users. It is found that self-supervised representations are useful as input features to non-intrusive prediction models, achieving competitive performance to more complex systems. A detailed analysis of the performance depending on Clarity Prediction Challenge 1 listeners and enhancement systems indicates that more data might be needed to allow generalisation to unknown systems and (hearing-impaired) individuals

George Close, Kris Hong, Thomas Hain et al.

There has been significant research effort developing neural-network-based predictors of SQ in recent years. While a primary objective has been to develop non-intrusive, i.e.~reference-free, metrics to assess the performance of SE systems, recent work has also investigated the direct inference of neural SQ predictors within the loss function of downstream speech tasks. To aid in the training of SQ predictors, several large datasets of audio with corresponding human labels of quality have been created. Recent work in this area has shown that speech representations derived from large unsupervised or semi-supervised foundational speech models are useful input feature representations for neural SQ prediction. In this work, a novel and robust SQ predictor is proposed based on feature representations extracted from an ASR model, found to be a powerful input feature for the SQ prediction task. The proposed system achieves higher correlation with human MOS ratings than recent approaches on all NISQA test sets and shows significantly better domain adaption compared to the commonly used DNSMOS metric.

William Ravenscroft, George Close, Kit Bower-Morris et al.

Large-scale in-the-wild speech datasets have become more prevalent in recent years due to increased interest in models that can learn useful features from unlabelled data for tasks such as speech recognition or synthesis. These datasets often contain undesirable features, such as multiple speakers, non-target languages, and music, which may impact model learning. The Whilter model is proposed as a multitask solution to identify these undesirable samples. Whilter uses a Whisper encoder with an attention-based classifier to solve five diverse classification problems at once. In addition, an annotated dataset is published for a subset of two popular in-the-wild corpora. Whilter achieves F1 scores above 85% and equal error rates of 6.5% to 7.8% for three of five subtasks, outperforming a state-of-the-art BEATs classifier on speech-specific classes, with a notable decrease in processing time compared to a combination of single-task alternatives.

William Ravenscroft, George Close, Stefan Goetze et al.

One solution to automatic speech recognition (ASR) of overlapping speakers is to separate speech and then perform ASR on the separated signals. Commonly, the separator produces artefacts which often degrade ASR performance. Addressing this issue typically requires reference transcriptions to jointly train the separation and ASR networks. This is often not viable for training on real-world in-domain audio where reference transcript information is not always available. This paper proposes a transcription-free method for joint training using only audio signals. The proposed method uses embedding differences of pre-trained ASR encoders as a loss with a proposed modification to permutation invariant training (PIT) called guided PIT (GPIT). The method achieves a 6.4% improvement in word error rate (WER) measures over a signal-level loss and also shows enhancement improvements in perceptual measures such as short-time objective intelligibility (STOI).

Rhiannon Mogridge, George Close, Robert Sutherland et al.

Neural networks have been successfully used for non-intrusive speech intelligibility prediction. Recently, the use of feature representations sourced from intermediate layers of pre-trained self-supervised and weakly-supervised models has been found to be particularly useful for this task. This work combines the use of Whisper ASR decoder layer representations as neural network input features with an exemplar-based, psychologically motivated model of human memory to predict human intelligibility ratings for hearing-aid users. Substantial performance improvement over an established intrusive HASPI baseline system is found, including on enhancement systems and listeners unseen in the training data, with a root mean squared error of 25.3 compared with the baseline of 28.7.