Multimodal Emotion Recognition baseline code for MERC2020
- This is baseline code for MERC2020
- This code is based on keras(2.2.4) and tensorflow(1.14).
- You need to prepare multimodal emotion dataset and speech recognition baseline code.
1. Get features using VGG face
CUDA_VISIBLE_DEVICES=0 python3 video_feature_extract.py
2. Load dataset
python3 load.py
3. Train model
CUDA_VISIBLE_DEVICES=0 python3 train.py
Performance
- Accuracy: 29.71% (valiation dataset)
Multimodal baseline model
Root path
cd modules/multimodal
Usage examples
1. Get bottleneck features from each single modal
CUDA_VISIBLE_DEVICES=0 python3 feature_extract.py
2. Train model
CUDA_VISIBLE_DEVICES=0 python3 train.py
Performance
- Accuracy: 52.56% (valiation dataset)
License
- The source code, pre-trained models, datasets are available under MERC2020 competition.