Revisiting Multimodal Emotion Recognition in Conversation from the Perspective of Graph Spectrum

Authors: Wei Ai, Fuchen Zhang, Yuntao Shou, Tao Meng, Haowen Chen, Keqin Li

AAAI 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experiments have proven the superiority of the GS-MCC architecture proposed in this paper on two benchmark data sets.
Researcher Affiliation Academia 1 College of Computer and Mathematics, Central South University of Forestry and Technology, 410004, China 2 College of Computer Science and Electronic Engineering, Hunan University, 410082, China 3 Department of Computer Science, State University of New York, 12561, USA
Pseudocode No The paper describes the methodology in text and mathematical formulas but does not include a clearly labeled pseudocode or algorithm block.
Open Source Code Yes Our code is publicly available at https://github.com/Fuchen Zhang/GS-MCC.
Open Datasets Yes In our experiments, we used two benchmark multimodal datasets IEMOCAP (Busso et al. 2008) and MELD (Poria et al. 2019), which are widely used in multimodal emotion recognition.
Dataset Splits Yes The optimal parameters of all models were obtained by performing parameter adjustment using the leave-one-out cross-validation method on the validation set.
Hardware Specification Yes All experiments are conducted using Python 3.8 and Py Torch 1.8 deep learning framework and performed on a single NVIDIA RTX 4090 24G GPU.
Software Dependencies Yes All experiments are conducted using Python 3.8 and Py Torch 1.8 deep learning framework
Experiment Setup Yes Our model is trained using Adam W with a learning rate of 1e5, cross-entropy as the loss function, and a batch size of 32.