Netflix vmaf: Perceptual video quality assessment based on multi-method fusion

Netflix has open-sourced VMAF, a perceptual video quality assessment tool that uses multi-method fusion to evaluate video quality. VMAF is a calculation of quality based on a combination of multiple algorithms, including SSIM, PSNR-HVS-M, and other models. The tool is designed to provide a more accurate assessment of video quality than traditional metrics like PSNR or SSIM, which can be misleading in certain situations. VMAF is currently available on GitHub and can be used to evaluate video quality on a variety of platforms, including web, mobile, and broadcast TV.