Awesome
MiE-X
Introduction
This is an implementation of MiE-X by Pytorch. MiE-X is a large-scale synthetic dataset that improves data-driven micro-expression methods. MiE-X introduces three types of effective Actions Units (AUs) that constitute trainable micro-expressions. This repository provides the implementation of acquiring these AUs and using these AUs to obtain MiE-X.
<!-- ## Overview Overview of computing three types of Action Units.  -->Dependencies
MiE-X uses the same libraries as GANimation
- python 3.7+
- pytorch 1.6+ & torchvision
- numpy
- matplotlib
- tqdm
- dlib
- face_recognition
- opencv-contrib-python
Datasets
We make generated images from VehicleX directly. We have performed domain adaptation (both content level and style level) from VehicleX to VeRi-776, VehicleID and CityFlow respectively. They can be used to augment real-world data. The adaptated images can be downloaded the tabel below.
| Variant | MiE-X (MEGC) | MiE-X (MMEW) | MiE-X (Oulu) |
|---|---|---|---|
| Access | Baidu(pwd:nz36),Google | Baidu(pwd:akjh),Google | Website |
Usage
Extract AUs by the OpenFace toolkit
python3 get_aus.py --persons_path PATH_TO_YOUR_VIDEOS
Simulate MiEs
use AU<sub>MiE</sub> to simulate
CUDA_VISIBLE_DEVICES=0 python3 simulate_realAU.py
use AU<sub>MaE</sub> to simulate
CUDA_VISIBLE_DEVICES=0 python3 simulate_ck.py
use AU<sub>exp</sub> to simulate
CUDA_VISIBLE_DEVICES=0 python3 simulate_data.py