/U-AnDi

Semantic Segmentation of Anomalous Diffusion Using Deep Convolutional Networks

Primary LanguagePython

U-AnDi

Supporting Codes of U-AnDi for data generation and model training

Semantic Segmentation of Anomalous Diffusion Using Deep Convolutional Networks

Xiang Qu, Yi Hu, Wenjie Cai, Yang Xu, Hu Ke, Guolong Zhu, and Zihan Huang

School of Physics and Electronics, Hunan University, Changsha 410082, China

E-mail: huangzih@hnu.edu.cn


Model

Note: For code files in this repository:

  • The suffix reg, cls, and exp correspond to subtask1, subtask2, and experimental data, respectively.

1. Data Generation

# Generation
python generate_data_reg.py --l 500 --N 1000000

# Pre-Processing for training
python pre_process_data_reg.py --l 500

l is the length, N is the number of trajectories.

The version of andi-datasets is 2.0.0 in subtask 1, and 0.0.5 in subtask 2 and experimental data.

2. Model Training

python train-U-AnDi-reg.py --l 500

Environment:

  • OS: Ubuntu 16.04
  • GPU: NVIDIA A100 40G
  • Python==3.7.4
  • PyTorch==1.12.1
  • cuda==11.6