IQA-PyTorch
This powerful toolbox leverages Python and PyTorch to offer efficient, GPU-accelerated image quality assessment. It supports a wide range of full reference and no-reference metrics, providing evaluations faster than those of traditional MATLAB scripts. The toolbox includes recent updates such as improved metrics and reduced GPU memory requirements. Users can install and utilize it via the command line or integrate it into larger projects with customizable settings and loss function integration. Benchmark performance comparisons and datasets are also provided for robust evaluations.