Search

Speech is a crucial source of personal information, and the risk of attackers using such information increases day by day. Speaker privacy protection is crucial, and various approaches have been proposed to hide the speaker’s identity. One approach is voice anonymization, which aims to safeguard speaker identity while maintaining speech content through techniques such as voice conversion or spectral feature alteration. The significance of voice anonymization has grown due to the necessity to protect personal information in applications such as voice assistants, authentication, and customer support. Building upon the S3PRL-VC toolkit and on pre-trained speech and speaker representation models, this paper introduces a feature disentanglement approach to improve the de-identification performance of the state-of-the-art anonymization approaches based on voice conversion. The proposed approach achieves state-of-the-art speaker de-identification and causes minimal impact on the intelligibility of the signal after conversion.

Abstract This paper tackles the front‐back disambiguity problem in speaker localization when the audio signals are captured by a symmetric microphone array. To this end, a deep neural network is proposed with an attention‐based mechanism designed to assign different weights to features obtained from individual microphones. For support, a real dataset with synchronized multichannel audio signals captured by a large linear microphone array is introduced, along with manual annotations. The experimental results demonstrate the effectiveness of the proposed method over the other approaches. In particular, more than 50% reduction in Equal Error Rate (EER) is achieved when comparing with the single‐channel case. The designed multi‐channel self‐attention mechanism also brings further improvements. The dataset and source code will be released.

Speech enhancement is a relevant component in many real-world applications such as hearing aid, mobile telecommunications and healthcare applications. In this paper, we investigate the Dilated Wave-U-Net model: a recently proposed end-to-end neural speech enhancement approach based on the Wave-U-Net architecture. We evaluate the performance of the model on two datasets: the public VCTK dataset, and a contaminated version of Librispeech. In particular, we experiment on using alternative losses based on L1 norm and on a combination of L1 and MSE losses. Results show that the Dilated Wave-U-Net architecture outperforms other state-of-the-art methods in terms of intelligibility and quality metrics on both datasets and that MSE loss is the most performing.

Robustness against background noise and reverberation is essential for many real-world speech-based applications. One way to achieve this robustness is to employ a speech enhancement front-end that, independently of the back-end, removes the environmental perturbations from the target speech signal. However, although the enhancement front-end typically increases the speech quality from an intelligibility perspective, it tends to introduce distortions which deteriorate the performance of subsequent processing modules. In this paper, we investigate strategies for jointly training neural models for both speech enhancement and the back-end, which optimize a combined loss function. In this way, the enhancement front-end is guided by the back-end to provide more effective enhancement. Differently from typical state-of-the-art approaches employing on spectral features or neural embeddings, we operate in the time domain, processing raw waveforms in both components. As application scenario we consider intent classification in noisy environments. In particular, the front-end speech enhancement module is based on Wave-U-Net while the intent classifier is implemented as a temporal convolutional network. Exhaustive experiments are reported on versions of the Fluent Speech Commands corpus contaminated with noises from the Microsoft Scalable Noisy Speech Dataset, shedding light and providing insight about the most promising training approaches.