-
About Engineering Applications of Artificial Intelligence (EAAI)
-
Nhut Minh Nguyen, Minh Trung Nguyen, Thanh Trung Nguyen, Phuong-Nam Tran, Nhat Truong Pham, Linh Le, Alice Othmani, Abdulmotaleb El Saddik and Duc Ngoc Minh Dang, “Enhancing multimodal emotion recognition with dynamic fuzzy membership and attention fusion”: This paper introduces FleSER, a new multimodal emotion recognition architecture that combines dynamic fuzzy membership with attention-based fusion. Unlike most existing SER models that apply fuzzy logic only at the decision level, FleSER uses a feature-level rule-based fuzzy mechanism to refine audio and text representations before fusion. Our design also incorporates self- and cross-modality attention, along with an α-interpolation fusion strategy, allowing the model to adaptively emphasize whichever modality (audio or text) is more informative in each context. FleSER achieves state-of-the-art performance across multiple benchmark datasets, with extensive ablation studies confirming the effectiveness of each architectural component. This work is a step forward in building more robust, adaptive, and accurate emotion recognition systems.