Variational Pathway Reasoning for EEG Emotion Recognition
Authors: Tong Zhang, Zhen Cui, Chunyan Xu, Wenming Zheng, Jian Yang2709-2716
AAAI 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experiments on EEG emotion recognition demonstrate that the proposed VPR is superior to those state-of-the-art methods, and could find some interesting pathways w.r.t. different emotions. |
| Researcher Affiliation | Academia | Tong Zhang,1 Zhen Cui,1 Chunyan Xu,1 Wenming Zheng,2 Jian Yang1 1Key Lab of Intelligent Perception and Systems for High-Dimensional Information of Ministry of Education, Jiangsu Key Lab of Image and Video Understanding for Social Security, School of Computer Science and Engineering, Nanjing University of Science and Technology, Nanjing, China. 2Research Center for Learning Science, School of Biological Science and Medical Engineering, Southeast University, Nanjing, China. {tong.zhang, zhen.cui, cyx}@njust.edu.cn, wenming zheng@seu.edu.cn, csjyang@njust.edu.cn |
| Pseudocode | No | No pseudocode or algorithm blocks were found. |
| Open Source Code | No | The paper does not provide any statement or link indicating the availability of open-source code for the described methodology. |
| Open Datasets | Yes | We evaluate the VPR by conducting experiments on two EEG emotion datasets named SJTU Emotion EEG Dataset (SEED) and multi-modal physiological emotion database (MPED). SEED (Zheng and Lu 2015) and MPED (Song et al. 2019) |
| Dataset Splits | Yes | For SEED, each subject conducts two times of experiments which yields totally 30 times of experiments. Following the cross session protocol in (Zheng and Lu 2015), the training and testing samples are respectively taken from different sessions of one experiment. As each time of experiment contains fifteen sessions, nine of them are used for training and the remaining six for testing. the architectures of the proposed VPR framework are set the same, which are determined by cross-validation on selected validation set (i.e., a part of training set). |
| Hardware Specification | No | The paper does not provide specific hardware details such as GPU/CPU models, memory, or processor types used for running experiments. |
| Software Dependencies | No | The paper does not list specific software dependencies with version numbers (e.g., Python, PyTorch, TensorFlow versions, or other libraries). |
| Experiment Setup | Yes | For random walk, four pathways are generated for each starting node (yielding 268 pathways in total) with the path length of 4, where every electrode is treated as the starting node in the process of random walk. Besides, the dimension of hidden state of LSTM for pathway embedding is traversed in the range of [8, 16, 32, 64], and finally set to 16. The value of λ in Eqn.12 is set to 0.5. In the training stage, we run the our VPR model for 20 epochs with a learning rate of 0.001 for tuning the network parameters, where the batch size is set to 64. |