• numpy
• html4vision
• shapely
• colour-science
• lensfunpy
• pytorch
• largestinteriorrectangle

Dataset Link

Download and unzip labels.zip to the code folder so that labels is at same level as various scripts.

images_png.zip and images_raw.zip are png format images and camera raw format images respectively. images_raw will be needed for shading computation.

In utils.py, change paths in ALGORITHM_PATHS to results output by each algorithm, and IMGS_PATH to path of MAW images.

In run_cmp.py, update names on line 30 to algorithms you want to evaluate. Similarly in texture_score_lpips.py, update names on line 750 to algothims you want to evaluate. (possible names are in utils.py starting at line 159.)

To evaluate all algorimths for albedo intensity, run:

python3 run_cmp.py --meta meta.csv --loss="si" --metric "mean" --type=metric

To evaluate all algorimths for albedo chromaticity, run:

python3 run_cmp.py --meta meta.csv --loss="per_si" --metric "deltae" --type=metric

To evaluate all algorimths for WHDR on MAW, run:

python3 run_cmp.py --meta meta.csv --type=whdr

To evaluate all algorithms for texture, run:

python3 texture_score_lpips.py maw --meta meta.csv --imgs_dir <MAW PNG IMG PATH>  --use_gpu

• With images_png and images_raw in the same folder, run python3 compute_shading.py meta.csv --imgs_dir <MAW PNG IMG PATH> to generate shading labels.
• In run_shading_cmp.py update names on line 54 to algorithms you want to evaluate.

To evaluate all algorimths for shading, run:

python3 run_shading_cmp.py --meta meta.csv --imgs_dir <MAW PNG IMG PATH>

Unfortunately, we have to remove 14 images from scene_0 as those images reveal personal credit cards. We will consider restore those images by blackout sensitive area.

The way number of scenes are counted in the dataset release is different from the paper. In the dataset release, some scene can contain multiple scenes as counted by the paper, as pictures in those scenes come from physically separate areas/different rooms. scene containing multiple scenes are listed below:

• <scene_0>: contains 2 scenes.
• <scene_2>: contains 3 scenes.
• <scene_2>: contains 2 scenes.
• <scene_31>: contains 4 scenes.
• <scene_34>: contains 2 scenes.