This alogorithm learns the desk tidying tasks with Asymmetric SAC+VAE.
The unit tasks for tidying the desk are closing the drawer and closing the laptop.
A random number of objects are placed on the desk so that it can be operated in a variety of environments.
An example of each task is as follows.
In this case, the algorithm uses RGB, depth, and segmentation images in the virtual environment.
The structure of Asymmetric SAC+VAE is as follows.
Learning proceeds through the following code.
python main.py --GUI GUI --cuda Download the urdf file for training from Asymmetric_SAC_VAE_urdf at the following link.
link for urdf: https://drive.google.com/drive/folders/1vb5gxkhQ9PLQNBXB_dwg47qoHiaw5EOL?usp=sharing