Joint Learning using Mixture-of-Expert-Based Representation for Enhanced Speech Generation and Robust Emotion Recognition

Speech emotion recognition (SER) plays a critical role in building emotion-aware speech systems, but its performance degrades significantly under noisy conditions. Although speech enhancement (SE) can improve robustness, it often introduces artifacts that obscure emotional cues and adds computational overhead to the pipeline. Multi-task learning (MTL) offers an alternative by jointly optimizing SE and SER tasks. However, conventional shared-backbone models frequently suffer from gradient interference and representational conflicts between tasks. To address these challenges, we propose the Sparse Mixture-of-Experts Representation Integration Technique (Sparse MERIT), a flexible MTL framework that applies frame-wise expert routing over self-supervised speech representations. Sparse MERIT incorporates task-specific gating networks that dynamically select from a shared pool of experts for each frame, enabling parameter-efficient and task-adaptive representation learning. Experiments on the MSP-Podcast corpus show that Sparse MERIT consistently outperforms baseline models on both SER and SE tasks. Under the most challenging condition of -5 dB signal-to-noise ratio (SNR), Sparse MERIT improves SER F1-macro by an average of 12.0% over a baseline relying on a SE pre-processing strategy, and by 3.4% over a naive MTL baseline, with statistical significance on unseen noise conditions. For SE, Sparse MERIT improves segmental SNR (SSNR) by 28.2% over the SE pre-processing baseline and by 20.0% over the naive MTL baseline. These results demonstrate that Sparse MERIT provides robust and generalizable performance for both emotion recognition and enhancement tasks in noisy environments.

Lessons Learnt: Revisit Key Training Strategies for Effective Speech Emotion Recognition in the Wild

In this study, we revisit key training strategies in machine learning often overlooked in favor of deeper architectures. Specifically, we explore balancing strategies, activation functions, and fine-tuning techniques to enhance speech emotion recognition (SER) in naturalistic conditions. Our findings show that simple modifications improve generalization with minimal architectural changes. Our multi-modal fusion model, integrating these optimizations, achieves a valence CCC of 0.6953, the best valence score in Task 2: Emotional Attribute Regression. Notably, fine-tuning RoBERTa and WavLM separately in a single-modality setting, followed by feature fusion without training the backbone extractor, yields the highest valence performance. Additionally, focal loss and activation functions significantly enhance performance without increasing complexity. These results suggest that refining core components, rather than deepening models, leads to more robust SER in-the-wild.

Is It Still Fair? Investigating Gender Fairness in Cross-Corpus Speech Emotion Recognition

Speech emotion recognition (SER) is a vital component in various everyday applications. Cross-corpus SER models are increasingly recognized for their ability to generalize performance. However, concerns arise regarding fairness across demographics in diverse corpora. Existing fairness research often focuses solely on corpus-specific fairness, neglecting its generalizability in cross-corpus scenarios. Our study focuses on this underexplored area, examining the gender fairness generalizability in cross-corpus SER scenarios. We emphasize that the performance of cross-corpus SER models and their fairness are two distinct considerations. Moreover, we propose the approach of a combined fairness adaptation mechanism to enhance gender fairness in the SER transfer learning tasks by addressing both source and target genders. Our findings bring one of the first insights into the generalizability of gender fairness in cross-corpus SER systems.

Mouth Articulation-Based Anchoring for Improved Cross-Corpus Speech Emotion Recognition

Cross-corpus speech emotion recognition (SER) plays a vital role in numerous practical applications. Traditional approaches to cross-corpus emotion transfer often concentrate on adapting acoustic features to align with different corpora, domains, or labels. However, acoustic features are inherently variable and error-prone due to factors like speaker differences, domain shifts, and recording conditions. To address these challenges, this study adopts a novel contrastive approach by focusing on emotion-specific articulatory gestures as the core elements for analysis. By shifting the emphasis on the more stable and consistent articulatory gestures, we aim to enhance emotion transfer learning in SER tasks. Our research leverages the CREMA-D and MSP-IMPROV corpora as benchmarks and it reveals valuable insights into the commonality and reliability of these articulatory gestures. The findings highlight mouth articulatory gesture potential as a better constraint for improving emotion recognition across different settings or domains.

Stimulus Modality Matters: Impact of Perceptual Evaluations from Different Modalities on Speech Emotion Recognition System Performance

Speech Emotion Recognition (SER) systems rely on speech input and emotional labels annotated by humans. However, various emotion databases collect perceptional evaluations in different ways. For instance, the IEMOCAP dataset uses video clips with sounds for annotators to provide their emotional perceptions. However, the most significant English emotion dataset, the MSP-PODCAST, only provides speech for raters to choose the emotional ratings. Nevertheless, using speech as input is the standard approach to training SER systems. Therefore, the open question is the emotional labels elicited by which scenarios are the most effective for training SER systems. We comprehensively compare the effectiveness of SER systems trained with labels elicited by different modality stimuli and evaluate the SER systems on various testing conditions. Also, we introduce an all-inclusive label that combines all labels elicited by various modalities. We show that using labels elicited by voice-only stimuli for training yields better performance on the test set, whereas labels elicited by voice-only stimuli.

DiffEVC: Any-to-Any Emotion Voice Conversion with Expressive Guidance

Emotional Voice Conversion (EVC) modifies speech emotion to enhance communication by amplifying positive cues and reducing negative ones. This complex task involves entangled factors like voice quality, speaker traits, and content. Traditional deep learning models like GANs and autoencoders have achieved some success in EVC by learning mappings or disentangling features but face challenges like instability and voice quality degradation. Diffusion models offer stable training and high-quality generation. We propose a diffusion-based EVC framework that disentangles emotion and speaker identity using mutual information loss and auxiliary models. An expressive guidance mechanism is introduced to improve emotion conversion while maintaining speaker traits. Experimental results demonstrate our approach's effectiveness for unseen speakers and emotions, achieving state-of-the-art performance in EVC tasks.

EMO-Codec: An In-Depth Look at Emotion Preservation capacity of Legacy and Neural Codec Models With Subjective and Objective Evaluations

The neural codec model reduces speech data transmission delay and serves as the foundational tokenizer for speech language models (speech LMs). Preserving emotional information in codecs is crucial for effective communication and context understanding. However, there is a lack of studies on emotion loss in existing codecs. This paper evaluates neural and legacy codecs using subjective and objective methods on emotion datasets like IEMOCAP. Our study identifies which codecs best preserve emotional information under various bitrate scenarios. We found that training codec models with both English and Chinese data had limited success in retaining emotional information in Chinese. Additionally, resynthesizing speech through these codecs degrades the performance of speech emotion recognition (SER), particularly for emotions like sadness, depression, fear, and disgust. Human listening tests confirmed these findings. This work guides future speech technology developments to ensure new codecs maintain the integrity of emotional information in speech.

EMO-Codec: A Depth Look at Emotion Preservation Capacity of Legacy and Neural Codec Models With Subjective and Objective Evaluations

The neural codec model reduces speech data transmission delay and serves as the foundational tokenizer for speech language models (speech LMs). Preserving emotional information in codecs is crucial for effective communication and context understanding. However, there is a lack of studies on emotion loss in existing codecs. This paper evaluates neural and legacy codecs using subjective and objective methods on emotion datasets like IEMOCAP. Our study identifies which codecs best preserve emotional information under various bitrate scenarios. We found that training codec models with both English and Chinese data had limited success in retaining emotional information in Chinese. Additionally, resynthesizing speech through these codecs degrades the performance of speech emotion recognition (SER), particularly for emotions like sadness, depression, fear, and disgust. Human listening tests confirmed these findings. This work guides future speech technology developments to ensure new codecs maintain the integrity of emotional information in speech.

Improving Robustness and Clinical Applicability of Respiratory Sound Classification via Audio Enhancement

Deep learning techniques have shown promising results in the automatic classification of respiratory sounds. However, accurately distinguishing these sounds in real-world noisy conditions poses challenges for clinical deployment. Additionally, predicting signals with only background noise could undermine user trust in the system. In this study, we propose an audio enhancement (AE) pipeline as a pre-processing step before respiratory sound classification, aiming to improve performance in noisy environments. Multiple experiments were conducted using different audio enhancement model structures, demonstrating improved classification performance compared to the baseline method of noise injection data augmentation. Specifically, the integration of the AE pipeline resulted in a 2.59% increase in the ICBHI classification score on the ICBHI respiratory sound dataset and a 2.51% improvement on our recently collected Formosa Archive of Breath Sounds (FABS) in multi-class noisy scenarios. Furthermore, a physician validation study assessed the clinical utility of our system. Quantitative analysis revealed enhancements in efficiency, diagnostic confidence, and trust during model-assisted diagnosis with our system compared to raw noisy recordings. Workflows integrating enhanced audio led to an 11.61% increase in diagnostic sensitivity and facilitated high-confidence diagnoses. Our findings demonstrate that incorporating an audio enhancement algorithm significantly enhances robustness and clinical utility.

A Layer-Anchoring Strategy for Enhancing Cross-Lingual Speech Emotion Recognition

Cross-lingual speech emotion recognition (SER) is important for a wide range of everyday applications. While recent SER research relies heavily on large pretrained models for emotion training, existing studies often concentrate solely on the final transformer layer of these models. However, given the task-specific nature and hierarchical architecture of these models, each transformer layer encapsulates different levels of information. Leveraging this hierarchical structure, our study focuses on the information embedded across different layers. Through an examination of layer feature similarity across different languages, we propose a novel strategy called a layer-anchoring mechanism to facilitate emotion transfer in cross-lingual SER tasks. Our approach is evaluated using two distinct language affective corpora (MSP-Podcast and BIIC-Podcast), achieving a best UAR performance of 60.21% on the BIIC-podcast corpus. The analysis uncovers interesting insights into the behavior of popular pretrained models.