This is code of RID-Noise.

Reproduce RID-Noise Results

Toy tasks

Please refer to the notebook ridnoise.ipynb to view experiments on three toy tasks.

Benchmarks

We provide the source code to reproduce the performance of RID-Noise in the Table 1. of the paper, including the implementation of all the three benchmark tasks.

To reproduce the results of the Kinematics task with x dependent noise, please run

python train_and_inference.py --method rid --task noisy_kine_x --device cuda:0

We provide a bash script to reproduce all the results, please execute

bash run.sh

Please cite our paper if you find this repository helpful in your work.

@inproceedings{rid_noise,
  author    = {Jia{-}Qi Yang and
               Ke{-}Bin Fan and
               Hao Ma and
               De{-}Chuan Zhan},
  title     = {RID-Noise: Towards Robust Inverse Design under Noisy Environments},
  booktitle = {Thirty-Sixth {AAAI} Conference on Artificial Intelligence, {AAAI}
               2022},
  year      = {2022}
}