FovVideoVDP

Description

Cambridge Rainbow Group’s foveated visible-difference predictor for video and images, with Python and Matlab implementations used as a perceptual quality metric beyond VMAF.

URL

• Visit resource

Related resources

• VMAF Reproducibility: Validating a Perceptual Practical Video Quality MetricPeer-reviewed IEEE paper on Netflix VMAF metric validation and reproducibility; authoritative reference for quality metrics in codec benchm…
• common_metrics_on_video_qualityPython runner computing FVD, PSNR, SSIM, and LPIPS on generated or predicted video sequences.
• objective-metricsCLI for image and video quality assessment covering learned metrics (MDTVSFA, FAST-VQA, MUSIQ) as well as VMAF.
• NoR-VDPNetDeep-learning no-reference approximation of HDR-VDP for estimating high-dynamic-range visible differences without a reference.
• vmaf-rsRust bindings for libvmaf so video pipelines can compute VMAF without shelling out to FFmpeg.
• MPEG-H Audio (Fraunhofer IIS)Technical overview from the primary developer of MPEG-H 3D Audio covering object-based next-generation audio for broadcast and streaming, i…
• ITU-T P.1204.3 reference softwareReference implementation of ITU-T P.1204.3, a bitstream-based no-reference video quality model for H.264, HEVC, and VP9 up to 4K, used to s…
• Evaluating Video Quality Metrics for Neural and Traditional Codecs (4K/UHD-1)Academic comparison of VMAF, AVQBits, and FasterVQA no-reference quality metrics on 4K content.
• vs-placeboVapourSynth plugin that uses libplacebo for GPU debanding, scaling, and color mapping.