LFASR-geometry

PyTorch implementation of AAAI 2020 paper: Learning Light Field Angular Super-Resolution via a Geometry-Aware Network.

[paper]

Requirements

• Python 3.6
• PyTorch 1.1
• Matlab (for training/test data generation)

Dataset

We provide MATLAB code for preparing the training and test data. Please first download light field datasets, and put them into corresponding folders in LFData.

Demo

To produce the results in the paper, run:

python test_pretrained.py --model_path ./pretrained_model/HCI_2x2-7x7.pth   --test_dataset HCI --data_path ./LFData/test_HCI.h5 --angular_out 7 --angular_in 2 --crop 1 --save_img 1

Training

To re-train the model, run:

python train.py --lr 1e-4 --step 500 --epi 1.0 --patch_size 96 --num_cp 10   --layer_num 4  --angular_out 7 --angular_in 2 --dataset HCI --dataset_path ./LFData/train_HCI.h5