Skip to content

Latest commit

 

History

History
103 lines (71 loc) · 4.88 KB

README.md

File metadata and controls

103 lines (71 loc) · 4.88 KB
Raw

PyTorch Image Quality Assessment

PIQA is a collection of PyTorch metrics for image quality assessment in various image processing tasks such as generation, denoising, super-resolution, interpolation, etc. It focuses on the efficiency, conciseness and understandability of its (sub-)modules, such that anyone can easily reuse and/or adapt them to its needs.

PIQA should be pronounced pika (like Pikachu ⚡️)

Installation

The piqa package is available on PyPI, which means it is installable via pip.

pip install piqa

Alternatively, if you need the latest features, you can install it from the repository.

pip install git+https://github.com/francois-rozet/piqa

Getting started

In piqa, each metric is associated to a class, child of torch.nn.Module, which has to be instantiated to evaluate the metric. All metrics are differentiable and support CPU and GPU (CUDA).

import torch
import piqa

# PSNR
x = torch.rand(5, 3, 256, 256)
y = torch.rand(5, 3, 256, 256)

psnr = piqa.PSNR()
l = psnr(x, y)

# SSIM
x = torch.rand(5, 3, 256, 256, requires_grad=True).cuda()
y = torch.rand(5, 3, 256, 256).cuda()

ssim = piqa.SSIM().cuda()
l = 1 - ssim(x, y)
l.backward()

Like torch.nn built-in components, these classes are based on functional definitions of the metrics, which are less user-friendly, but more versatile.

from piqa.ssim import ssim
from piqa.utils.functional import gaussian_kernel

kernel = gaussian_kernel(11, sigma=1.5).repeat(3, 1, 1)
ss, cs = ssim(x, y, kernel=kernel)

For more information, check out the documentation at piqa.readthedocs.io.

Available metrics

Class Range Objective Year Metric
TV [0, ∞] / 1937 Total Variation
PSNR [0, ∞] max / Peak Signal-to-Noise Ratio
SSIM [0, 1] max 2004 Structural Similarity
MS_SSIM [0, 1] max 2004 Multi-Scale Structural Similarity
LPIPS [0, ∞] min 2018 Learned Perceptual Image Patch Similarity
GMSD [0, ∞] min 2013 Gradient Magnitude Similarity Deviation
MS_GMSD [0, ∞] min 2017 Multi-Scale Gradient Magnitude Similarity Deviation
MDSI [0, ∞] min 2016 Mean Deviation Similarity Index
HaarPSI [0, 1] max 2018 Haar Perceptual Similarity Index
VSI [0, 1] max 2014 Visual Saliency-based Index
FSIM [0, 1] max 2011 Feature Similarity
FID [0, ∞] min 2017 Fréchet Inception Distance

Tracing

All metrics of piqa support PyTorch's tracing, which optimizes their execution, especially on GPU.

ssim = piqa.SSIM().cuda()
ssim_traced = torch.jit.trace(ssim, (x, y))

l = 1 - ssim_traced(x, y)  # should be faster ¯\_(ツ)_/¯

Assert

PIQA uses type assertions to raise meaningful messages when a metric doesn't receive an input of the expected type. This feature eases a lot early prototyping and debugging, but it might hurt a little the performances. If you need the absolute best performances, the assertions can be disabled with the Python flag -O. For example,

python -O your_awesome_code_using_piqa.py

Alternatively, you can disable PIQA's type assertions within your code with

piqa.utils.set_debug(False)

Contributing

If you have a question, an issue or would like to contribute, please read our contributing guidelines.