Joshua Reiss

Bimodal speech emotion recognition is challenging due to the difficulty of capturing subtle emotional cues and effectively modeling complex interactions between audio and text. Existing methods often overlook the importance of modality connections and struggle to filter noisy or misleading cross-modal signals. To address these issues, we propose a novel Bimodal Connection Attention Fusion (BCAF) framework that explicitly models modality-specific representations and their interactions. BCAF includes three key modules: an interactive connection network that learns modality connections through an encoderdecoder structure; a bimodal attention network that captures intra- and inter-modal dependencies; and a correlative attention network that filters cross-modal noise to improve robustness. Extensive evaluations across benchmark datasets show that BCAF achieves superior performance over existing methods, highlighting its strength in modeling detailed modality connections and interactions for emotion recognition.

Neural Audio Synthesis is dedicated to generating sound through generative neural networks. Sound effects are defined as auditory elements that complement a specific scene (in cinema, fiction, or videogames), support a storyline, enhance a fictional environment, or improve the perceived plausibility and presence (including Virtual Reality) without being music or dialog. This manuscript presents a quantitative literature review of the literature that intersects these two domains: the neural generation of sound effects. By leveraging large language models, we performed an extensive and systematic survey of the major scientific repositories, filtering the most relevant articles to ensure a thorough analysis. Our study examines various generation paradigms employed in sound synthesis, the specific types of sound effects created, the datasets used, and the evaluation metrics considered. Furthermore, we provide a forward-looking discussion on the evolution of this field towards multimodal approaches, where sound generation might integrate with other sensory modalities. All supporting materials and code are available online.