Investigating Effective Speaker Property Privacy Protection in Federated Learning for Speech Emotion Recognition

Federated Learning (FL) is a privacy-preserving approach that allows servers to aggregate distributed models transmitted from local clients rather than training on user data. More recently, FL has been applied to Speech Emotion Recognition (SER) for secure human-computer interaction applications. Recent research has found that FL is still vulnerable to inference attacks. To this end, this paper focuses on investigating the security of FL for SER concerning property inference attacks. We propose a novel method to protect the property information in speech data by decomposing various properties in the sound and adding perturbations to these properties. Our experiments show that the proposed method offers better privacy-utility trade-offs than existing methods. The trade-offs enable more effective attack prevention while maintaining similar FL utility levels. This work can guide future work on privacy protection methods in speech processing.

Optimizing Cox Models with Stochastic Gradient Descent: Theoretical Foundations and Practical Guidances

Optimizing Cox regression and its neural network variants poses substantial computational challenges in large-scale studies. Stochastic gradient descent (SGD), known for its scalability in model optimization, has recently been adapted to optimize Cox models. Unlike its conventional application, which typically targets a sum of independent individual loss, SGD for Cox models updates parameters based on the partial likelihood of a subset of data. Despite its empirical success, the theoretical foundation for optimizing Cox partial likelihood with SGD is largely underexplored. In this work, we demonstrate that the SGD estimator targets an objective function that is batch-size-dependent. We establish that the SGD estimator for the Cox neural network (Cox-NN) is consistent and achieves the optimal minimax convergence rate up to a polylogarithmic factor. For Cox regression, we further prove the $\sqrt{n}$-consistency and asymptotic normality of the SGD estimator, with variance depending on the batch size. Furthermore, we quantify the impact of batch size on Cox-NN training and its effect on the SGD estimator's asymptotic efficiency in Cox regression. These findings are validated by extensive numerical experiments and provide guidance for selecting batch sizes in SGD applications. Finally, we demonstrate the effectiveness of SGD in a real-world application where GD is unfeasible due to the large scale of data.

Speech Emotion Recognition under Resource Constraints with Data Distillation

Speech emotion recognition (SER) plays a crucial role in human-computer interaction. The emergence of edge devices in the Internet of Things (IoT) presents challenges in constructing intricate deep learning models due to constraints in memory and computational resources. Moreover, emotional speech data often contains private information, raising concerns about privacy leakage during the deployment of SER models. To address these challenges, we propose a data distillation framework to facilitate efficient development of SER models in IoT applications using a synthesised, smaller, and distilled dataset. Our experiments demonstrate that the distilled dataset can be effectively utilised to train SER models with fixed initialisation, achieving performances comparable to those developed using the original full emotional speech dataset.

Diff-ETS: Learning a Diffusion Probabilistic Model for Electromyography-to-Speech Conversion

Electromyography-to-Speech (ETS) conversion has demonstrated its potential for silent speech interfaces by generating audible speech from Electromyography (EMG) signals during silent articulations. ETS models usually consist of an EMG encoder which converts EMG signals to acoustic speech features, and a vocoder which then synthesises the speech signals. Due to an inadequate amount of available data and noisy signals, the synthesised speech often exhibits a low level of naturalness. In this work, we propose Diff-ETS, an ETS model which uses a score-based diffusion probabilistic model to enhance the naturalness of synthesised speech. The diffusion model is applied to improve the quality of the acoustic features predicted by an EMG encoder. In our experiments, we evaluated fine-tuning the diffusion model on predictions of a pre-trained EMG encoder, and training both models in an end-to-end fashion. We compared Diff-ETS with a baseline ETS model without diffusion using objective metrics and a listening test. The results indicated the proposed Diff-ETS significantly improved speech naturalness over the baseline.

A Survey of Privacy-Preserving Model Explanations: Privacy Risks, Attacks, and Countermeasures

As the adoption of explainable AI (XAI) continues to expand, the urgency to address its privacy implications intensifies. Despite a growing corpus of research in AI privacy and explainability, there is little attention on privacy-preserving model explanations. This article presents the first thorough survey about privacy attacks on model explanations and their countermeasures. Our contribution to this field comprises a thorough analysis of research papers with a connected taxonomy that facilitates the categorisation of privacy attacks and countermeasures based on the targeted explanations. This work also includes an initial investigation into the causes of privacy leaks. Finally, we discuss unresolved issues and prospective research directions uncovered in our analysis. This survey aims to be a valuable resource for the research community and offers clear insights for those new to this domain. To support ongoing research, we have established an online resource repository, which will be continuously updated with new and relevant findings. Interested readers are encouraged to access our repository at https://github.com/tamlhp/awesome-privex.

STAA-Net: A Sparse and Transferable Adversarial Attack for Speech Emotion Recognition

Speech contains rich information on the emotions of humans, and Speech Emotion Recognition (SER) has been an important topic in the area of human-computer interaction. The robustness of SER models is crucial, particularly in privacy-sensitive and reliability-demanding domains like private healthcare. Recently, the vulnerability of deep neural networks in the audio domain to adversarial attacks has become a popular area of research. However, prior works on adversarial attacks in the audio domain primarily rely on iterative gradient-based techniques, which are time-consuming and prone to overfitting the specific threat model. Furthermore, the exploration of sparse perturbations, which have the potential for better stealthiness, remains limited in the audio domain. To address these challenges, we propose a generator-based attack method to generate sparse and transferable adversarial examples to deceive SER models in an end-to-end and efficient manner. We evaluate our method on two widely-used SER datasets, Database of Elicited Mood in Speech (DEMoS) and Interactive Emotional dyadic MOtion CAPture (IEMOCAP), and demonstrate its ability to generate successful sparse adversarial examples in an efficient manner. Moreover, our generated adversarial examples exhibit model-agnostic transferability, enabling effective adversarial attacks on advanced victim models.

A Comprehensive Survey on Heart Sound Analysis in the Deep Learning Era

Heart sound auscultation has been demonstrated to be beneficial in clinical usage for early screening of cardiovascular diseases. Due to the high requirement of well-trained professionals for auscultation, automatic auscultation benefiting from signal processing and machine learning can help auxiliary diagnosis and reduce the burdens of training professional clinicians. Nevertheless, classic machine learning is limited to performance improvement in the era of big data. Deep learning has achieved better performance than classic machine learning in many research fields, as it employs more complex model architectures with stronger capability of extracting effective representations. Deep learning has been successfully applied to heart sound analysis in the past years. As most review works about heart sound analysis were given before 2017, the present survey is the first to work on a comprehensive overview to summarise papers on heart sound analysis with deep learning in the past six years 2017--2022. We introduce both classic machine learning and deep learning for comparison, and further offer insights about the advances and future research directions in deep learning for heart sound analysis.

Cutting Through the Noise: An Empirical Comparison of Psychoacoustic and Envelope-based Features for Machinery Fault Detection

Acoustic-based fault detection has a high potential to monitor the health condition of mechanical parts. However, the background noise of an industrial environment may negatively influence the performance of fault detection. Limited attention has been paid to improving the robustness of fault detection against industrial environmental noise. Therefore, we present the Lenze production background-noise (LPBN) real-world dataset and an automated and noise-robust auditory inspection (ARAI) system for the end-of-line inspection of geared motors. An acoustic array is used to acquire data from motors with a minor fault, major fault, or which are healthy. A benchmark is provided to compare the psychoacoustic features with different types of envelope features based on expert knowledge of the gearbox. To the best of our knowledge, we are the first to apply time-varying psychoacoustic features for fault detection. We train a state-of-the-art one-class-classifier, on samples from healthy motors and separate the faulty ones for fault detection using a threshold. The best-performing approaches achieve an area under curve of 0.87 (logarithm envelope), 0.86 (time-varying psychoacoustics), and 0.91 (combination of both).

Towards Efficient ECG-based Atrial Fibrillation Detection via Parameterised Hypercomplex Neural Networks

Atrial fibrillation (AF) is the most common cardiac arrhythmia and associated with a higher risk for serious conditions like stroke. Long-term recording of the electrocardiogram (ECG) with wearable devices embedded with an automatic and timely evaluation of AF helps to avoid life-threatening situations. However, the use of a deep neural network for auto-analysis of ECG on wearable devices is limited by its complexity. In this work, we propose lightweight convolutional neural networks (CNNs) for AF detection inspired by the recently proposed parameterised hypercomplex (PH) neural networks. Specifically, the convolutional and fully-connected layers of a real-valued CNN are replaced by PH convolutions and multiplications, respectively. PH layers are flexible to operate in any channel dimension n and able to capture inter-channel relations. We evaluate PH-CNNs on publicly available databases of dynamic and in-hospital ECG recordings and show comparable performance to corresponding real-valued CNNs while using approx. $1/n$ model parameters.

Knowledge Transfer For On-Device Speech Emotion Recognition with Neural Structured Learning

Speech emotion recognition (SER) has been a popular research topic in human-computer interaction (HCI). As edge devices are rapidly springing up, applying SER to edge devices is promising for a huge number of HCI applications. Although deep learning has been investigated to improve the performance of SER by training complex models, the memory space and computational capability of edge devices represents a constraint for embedding deep learning models. We propose a neural structured learning (NSL) framework through building synthesized graphs. An SER model is trained on a source dataset and used to build graphs on a target dataset. A lightweight model is then trained with the speech samples and graphs together as the input. Our experiments demonstrate that training a lightweight SER model on the target dataset with speech samples and graphs can not only produce small SER models, but also enhance the model performance over models with speech samples only.