👁️ 🖼️ 🔥PyTorch Toolbox for Image Quality Assessment, including LPIPS, FID, NIQE, NRQM(Ma), MUSIQ, TOPIQ, NIMA, DBCNN, BRISQUE, PI and more...
OTHER License
Bot releases are hidden (Show)
You can now use pyiqa
in terminal like this:
# list all available metrics
pyiqa -ls
# test with default settings
pyiqa [metric_name(s)] --target [image_path or dir] --ref [image_path or dir]
UNIQUE
by @zwx8981 at https://github.com/chaofengc/IQA-PyTorch/pull/146
Many thanks to the valuable contributions 🤗 !
Full Changelog: https://github.com/chaofengc/IQA-PyTorch/compare/v0.1.10...v0.1.11
Published by chaofengc 10 months ago
liqe
liqe_mix
by @zwx8981nima-koniq
and nima-spaq
bd385c7d6cfe24c9be8f9d92c1919f98c49d9538Full Changelog: https://github.com/chaofengc/IQA-PyTorch/compare/v0.1.8...v0.1.10
Published by chaofengc about 1 year ago
mad
option test_y_channel
to be false by default cf84ffba7031d2eafe6aabe8193fa8672a9b5638true
in inference model to force read RGB image ad93cd42455ef0aa7285ca980ee959b11111d728topiq_fr
, topiq_nr
, topiq_nr-face
, topiq_iaa
etc. introduced by our paper TOPIQ
st-lpips
by @abhijay9 in https://github.com/chaofengc/IQA-PyTorch/pull/93. Many thanks ❤️ !laion_aes
introduced by LAION-Aesthetics_Predictor
maniqa
7ee5ea6eaf78c0c3cb5b3b587df5fc75ae886783brisque
to force gray input 4e2c707f4d43792a54399eccb558f1672e69ac70Full Changelog: https://github.com/chaofengc/IQA-PyTorch/compare/v0.1.7...v0.1.8
Published by chaofengc over 1 year ago
as_loss=True
e027618de46a7f43238a210e35afd144eae07e47niqe
with gray scale input 212ecefebd466cac019a1678e1b3a4e0378face0uranker
b142d1c3cca730e128f1b0935b2b96cd7cc7aec5maniqa-koniq, maniqa-kadid
fe95923f9c48188c65666930048597b45c9046declipscore
for image-caption matching ecb3e5e58b47212655aa1e87d7faf162b5087788entropy
to calculate gray scale image entropy like matlab 5f6d4fbc79d14d56b29c0a0c9c84baa8d33e78c5Published by chaofengc over 1 year ago
We have identified and resolved a critical bug in the NRQM calculation within our toolbox. The issue stemmed from the SSIM (Structural Similarity Index) function, where only the structure similarity score was being utilized. We apologize for any inconvenience this may have caused.
As a result of this bug, the PI, which is determined by the formula PI = 0.5 * (10 - NRQM + NIQE), was also affected.
With this hotfix, we have:
We strongly recommend using the latest release to benefit from these crucial fixes.
If you encounter any issues or have further questions, please don't hesitate to reach out to our support team. Thank you for your understanding and continued support.
Published by chaofengc over 1 year ago
psnry
for y colorspace; ssimc
for RGB ssim
Published by chaofengc almost 2 years ago
metric_mode
option for list_models
as_loss
option to enable gradient backpropagation for metric. Default False
.epoch
instead of iteration
in lr schedulerclean_state_dict
before loading pretrain modelPublished by chaofengc over 2 years ago
as_loss
option to enable gradient backpropagation for metric. Default False
.filter2
function to matlab utilsreduction
option to EMDLosscrop_border
option to PSNR, SSIMPublished by chaofengc over 2 years ago
nanmean
, nancov
, blockproc
, fspecial
.Published by chaofengc over 2 years ago
We change the default musiq
weights from musiq-ava
to musiq-koniq
because it is more robust according to NR benchmark results
Lambda
transform in dataset to enable distributed trainingPublished by chaofengc over 2 years ago
lower_better
flag. This indicates whether a lower metric score is better.Published by chaofengc over 2 years ago
Published by chaofengc over 2 years ago
First experimental release version of pyiqa tools 😃 . It supports
pip install pyiqa
pyiqa.list_models()
Hope this will help your research and project. We will add more features and pretrained models.
And welcome contribute, and report bugs ! 🍻
Published by chaofengc over 2 years ago
This release contains
.csv
files: meta information of different datasets.pkl
files: train/split of different datasets