Depression diagnosis from patient interviews using multimodal machine learning

Background: Depression is a major public health concern, affecting an estimated five percent of the global population. Early and accurate diagnosis is essential to initiate effective treatment, yet recognition remains challenging in many clinical contexts. Speech, language, and behavioral cues collected during patient interviews may provide objective markers that support clinical assessment. Methods: We developed a diagnostic approach that integrates features derived from patient interviews, including speech patterns, linguistic characteristics, and structured clinical information. Separate models were trained for each modality and subsequently combined through multimodal fusion to reflect the complexity of real-world psychiatric assessment. Model validity was assessed with established performance metrics, and further evaluated using calibration and decision-analytic approaches to estimate potential clinical utility. Results: The multimodal model achieved superior diagnostic accuracy compared to single-modality models, with an AUROC of 0.88 and an F1-score of 0.75. Importantly, the fused model demonstrated good calibration and offered higher net clinical benefit compared to baseline strategies, highlighting its potential to assist clinicians in identifying patients with depression more reliably. Conclusion: Multimodal analysis of patient interviews using machine learning may serve as a valuable adjunct to psychiatric evaluation. By combining speech, language, and clinical features, this approach provides a robust framework that could enhance early detection of depressive disorders and support evidence-based decision-making in mental healthcare.

A Framework for Robust Speaker Verification in Highly Noisy Environments Leveraging Both Noisy and Enhanced Audio

Recent advancements in speaker verification techniques show promise, but their performance often deteriorates significantly in challenging acoustic environments. Although speech enhancement methods can improve perceived audio quality, they may unintentionally distort speaker-specific information, which can affect verification accuracy. This problem has become more noticeable with the increasing use of generative deep neural networks (DNNs) for speech enhancement. While these networks can produce intelligible speech even in conditions of very low signal-to-noise ratio (SNR), they may also severely alter distinctive speaker characteristics. To tackle this issue, we propose a novel neural network framework that effectively combines speaker embeddings extracted from both noisy and enhanced speech using a Siamese architecture. This architecture allows us to leverage complementary information from both sources, enhancing the robustness of speaker verification under severe noise conditions. Our framework is lightweight and agnostic to specific speaker verification and speech enhancement techniques, enabling the use of a wide range of state-of-the-art solutions without modification. Experimental results demonstrate the superior performance of our proposed framework.

SegReConcat: A Data Augmentation Method for Voice Anonymization Attack

Anonymization of voice seeks to conceal the identity of the speaker while maintaining the utility of speech data. However, residual speaker cues often persist, which pose privacy risks. We propose SegReConcat, a data augmentation method for attacker-side enhancement of automatic speaker verification systems. SegReConcat segments anonymized speech at the word level, rearranges segments using random or similarity-based strategies to disrupt long-term contextual cues, and concatenates them with the original utterance, allowing an attacker to learn source speaker traits from multiple perspectives. The proposed method has been evaluated in the VoicePrivacy Attacker Challenge 2024 framework across seven anonymization systems, SegReConcat improves de-anonymization on five out of seven systems.

On the Application of Diffusion Models for Simultaneous Denoising and Dereverberation

Diffusion models have been shown to achieve natural-sounding enhancement of speech degraded by noise or reverberation. However, their simultaneous denoising and dereverberation capability has so far not been studied much, although this is arguably the most common scenario in a practical application. In this work, we investigate different approaches to enhance noisy and/or reverberant speech. We examine the cascaded application of models, each trained on only one of the distortions, and compare it with a single model, trained either solely on data that is both noisy and reverberated, or trained on data comprising subsets of purely noisy, of purely reverberated, and of noisy reverberant speech. Tests are performed both on artificially generated and real recordings of noisy and/or reverberant data. The results show that, when using the cascade of models, satisfactory results are only achieved if they are applied in the order of the dominating distortion. If only a single model is desired that can operate on all distortion scenarios, the best compromise appears to be a model trained on the aforementioned three subsets of degraded speech data.

Audio-Visual Feature Synchronization for Robust Speech Enhancement in Hearing Aids

Audio-visual feature synchronization for real-time speech enhancement in hearing aids represents a progressive approach to improving speech intelligibility and user experience, particularly in strong noisy backgrounds. This approach integrates auditory signals with visual cues, utilizing the complementary description of these modalities to improve speech intelligibility. Audio-visual feature synchronization for real-time SE in hearing aids can be further optimized using an efficient feature alignment module. In this study, a lightweight cross-attentional model learns robust audio-visual representations by exploiting large-scale data and simple architecture. By incorporating the lightweight cross-attentional model in an AVSE framework, the neural system dynamically emphasizes critical features across audio and visual modalities, enabling defined synchronization and improved speech intelligibility. The proposed AVSE model not only ensures high performance in noise suppression and feature alignment but also achieves real-time processing with minimal latency (36ms) and energy consumption. Evaluations on the AVSEC3 dataset show the efficiency of the model, achieving significant gains over baselines in perceptual quality (PESQ:0.52), intelligibility (STOI:19\%), and fidelity (SI-SDR:10.10dB).

Attention2Probability: Attention-Driven Terminology Probability Estimation for Robust Speech-to-Text System

Recent advances in speech large language models (SLMs) have improved speech recognition and translation in general domains, but accurately generating domain-specific terms or neologisms remains challenging. To address this, we propose Attention2Probability: attention-driven terminology probability estimation for robust speech-to-text system, which is lightweight, flexible, and accurate. Attention2Probability converts cross-attention weights between speech and terminology into presence probabilities, and it further employs curriculum learning to enhance retrieval accuracy. Furthermore, to tackle the lack of data for speech-to-text tasks with terminology intervention, we create and release a new speech dataset with terminology to support future research in this area. Experimental results show that Attention2Probability significantly outperforms the VectorDB method on our test set. Specifically, its maximum recall rates reach 92.57% for Chinese and 86.83% for English. This high recall is achieved with a latency of only 8.71ms per query. Intervening in SLMs' recognition and translation tasks using Attention2Probability-retrieved terms improves terminology accuracy by 6-17%, while revealing that the current utilization of terminology by SLMs has limitations.

Interpolating Speaker Identities in Embedding Space for Data Expansion

The success of deep learning-based speaker verification systems is largely attributed to access to large-scale and diverse speaker identity data. However, collecting data from more identities is expensive, challenging, and often limited by privacy concerns. To address this limitation, we propose INSIDE (Interpolating Speaker Identities in Embedding Space), a novel data expansion method that synthesizes new speaker identities by interpolating between existing speaker embeddings. Specifically, we select pairs of nearby speaker embeddings from a pretrained speaker embedding space and compute intermediate embeddings using spherical linear interpolation. These interpolated embeddings are then fed to a text-to-speech system to generate corresponding speech waveforms. The resulting data is combined with the original dataset to train downstream models. Experiments show that models trained with INSIDE-expanded data outperform those trained only on real data, achieving 3.06\% to 5.24\% relative improvements. While INSIDE is primarily designed for speaker verification, we also validate its effectiveness on gender classification, where it yields a 13.44\% relative improvement. Moreover, INSIDE is compatible with other augmentation techniques and can serve as a flexible, scalable addition to existing training pipelines.

Improving Noise Robust Audio-Visual Speech Recognition via Router-Gated Cross-Modal Feature Fusion

Robust audio-visual speech recognition (AVSR) in noisy environments remains challenging, as existing systems struggle to estimate audio reliability and dynamically adjust modality reliance. We propose router-gated cross-modal feature fusion, a novel AVSR framework that adaptively reweights audio and visual features based on token-level acoustic corruption scores. Using an audio-visual feature fusion-based router, our method down-weights unreliable audio tokens and reinforces visual cues through gated cross-attention in each decoder layer. This enables the model to pivot toward the visual modality when audio quality deteriorates. Experiments on LRS3 demonstrate that our approach achieves an 16.51-42.67% relative reduction in word error rate compared to AV-HuBERT. Ablation studies confirm that both the router and gating mechanism contribute to improved robustness under real-world acoustic noise.

Emotion Omni: Enabling Empathetic Speech Response Generation through Large Language Models

With the development of speech large language models (speech LLMs), users can now interact directly with assistants via speech. However, most existing models simply convert the response content into speech without fully understanding the rich emotional and paralinguistic cues embedded in the user's query. In many cases, the same sentence can have different meanings depending on the emotional expression. Furthermore, emotional understanding is essential for improving user experience in human-machine interaction. Currently, most speech LLMs with empathetic capabilities are trained on massive datasets. This approach requires vast amounts of data and significant computational resources. Therefore, a key challenge lies in how to develop a speech LLM capable of generating empathetic responses with limited data and without the need for large-scale training. To address this challenge, we propose Emotion Omni, a novel model architecture designed to understand the emotional content of user speech input and generate empathetic speech responses. Additionally, we developed a data generation pipeline based on an open-source TTS framework to construct a 200k emotional dialogue dataset, which supports the construction of an empathetic speech assistant. The demos are available at https://w311411.github.io/omni_demo/

MDD: a Mask Diffusion Detector to Protect Speaker Verification Systems from Adversarial Perturbations

Speaker verification systems are increasingly deployed in security-sensitive applications but remain highly vulnerable to adversarial perturbations. In this work, we propose the Mask Diffusion Detector (MDD), a novel adversarial detection and purification framework based on a \textit{text-conditioned masked diffusion model}. During training, MDD applies partial masking to Mel-spectrograms and progressively adds noise through a forward diffusion process, simulating the degradation of clean speech features. A reverse process then reconstructs the clean representation conditioned on the input transcription. Unlike prior approaches, MDD does not require adversarial examples or large-scale pretraining. Experimental results show that MDD achieves strong adversarial detection performance and outperforms prior state-of-the-art methods, including both diffusion-based and neural codec-based approaches. Furthermore, MDD effectively purifies adversarially-manipulated speech, restoring speaker verification performance to levels close to those observed under clean conditions. These findings demonstrate the potential of diffusion-based masking strategies for secure and reliable speaker verification systems.