BEGIN:VCALENDAR VERSION:2.0 PRODID:-//swoogo.com//NONSGML kigkonsult.se iCalcreator 2.41.90// CALSCALE:GREGORIAN UID:61623334-3363-4066-b663-643531336363 BEGIN:VEVENT UID:ba7f44ad89965efd1c9f92e12d7a7975a2ffbf21@swoogo.com DTSTAMP:20260813T001029Z DESCRIPTION:High Dynamic Range (HDR) and Wide Color Gamut (WCG) content is now mainstream across content creation\, with content playback supported o n millions of devices. Thus\, having a reliable way of evaluating HDR syst ems is essential. One common way of measuring the quality of an HDR system is measuring the color errors introduced along the imaging pipeline. Unfo rtunately\, performance of color difference metrics has mostly been evalua ted on databases composed of simple test patches\, as opposed to natural i magery. Some key differences between test patches and natural imagery is t hat test patches typically involve lower frequencies and non-contiguous re gions\, while the natural imagery has much higher frequencies\, masking du e to texture\, as well contiguous color region effects and gradients. \n\n Thus\, we evaluate several color difference metrics on five publicly avail able HDR databases consisting of natural images and subjective scores. The different databases focus on differing distortions and the aggregation of these cover a wide variety of both luminance and chromatic distortions. T here’re lower frequency distortions resulting from tone-mapping and gamut mapping operations. In addition\, there’re higher frequency distortions re sulting from compression artifacts by the various compression schemes such as JPEG\, JPEG-XT\, JPEG2000\, and HEVC. While perceptually dominated by luminance distortions\, these also contain physical chromatic distortions due to chromatic subsampling and different processes acting on Y\, Cr and Cb signals. Since it is desirable to evaluate as many images as possible\, a total of 64 source images and a total of 672 distorted images were eval uated by 94 observers across all databases. \n\nThe color difference metri cs we analyze include CIE94 and CIEDE00 metrics based on the CIE L*a*b* co lor space. In addition\, we analyzed the newer ones derived for HDR applic ations: DEITP based on the ICTCP color space\, and DEz based on the Jzazbz color space. Since it’s generally agreed that color quality doesn’t chang e significantly when motion is added\, we used still image databases for e valuation. \n\nTo quantify the performance\, we use four standard performa nce evaluation procedures – Root mean square error\, Pearson linear correl ation coefficient\, Spearman rank-order correlation coefficient and Outlie r ratio. The color spaces derived for HDR were the best performers across the different databases\, but neither of those two metrics performed the b est for every database. These databases have different experimental condit ions and display specifications. Analysis is currently underway to underst and why the two different HDR color space metrics performed best in terms of these differing conditions. DTSTART:20191023T213000Z DTEND:20191023T220000Z LAST-MODIFIED:20260813T001029Z LOCATION:Sacramento Room SEQUENCE:0 STATUS:CONFIRMED SUMMARY:HDR and WCG Image Quality Assessment Using Color Difference Metrics TRANSP:OPAQUE X-ALT-DESC;FMTTYPE=text/html:High Dynamic Range (HDR) and Wide Color Gamut (WCG) content is now mainstream across content creation\, with content pla yback supported on millions of devices. Thus\, having a reliable way of ev aluating HDR systems is essential. One common way of measuring the quality of an HDR system is measuring the color errors introduced along the imagi ng pipeline. Unfortunately\, performance of color difference metrics has m ostly been evaluated on databases composed of simple test patches\, as opp osed to natural imagery. Some key differences between test patches and nat ural imagery is that test patches typically involve lower frequencies and non-contiguous regions\, while the natural imagery has much higher frequen cies\, masking due to texture\, as well contiguous color region effects an d gradients.

\nThus\, we evaluate several color difference met rics on five publicly available HDR databases consisting of natural images and subjective scores. The different databases focus on differing distort ions and the aggregation of these cover a wide variety of both luminance a nd chromatic distortions. There’re lower frequency distortions resulting from tone-mapping and gamut mapping operations. In addition\, there’re hig her frequency distortions resulting from compression artifacts by the vari ous compression schemes such as JPEG\, JPEG-XT\, JPEG2000\, and HEVC. Whil e perceptually dominated by luminance distortions\, these also contain phy sical chromatic distortions due to chromatic subsampling and different pro cesses acting on Y\, Cr and Cb signals. Since it is desirable to evaluate as many images as possible\, a total of 64 source images and a total of 67 2 distorted images were evaluated by 94 observers across all databases.
\n The color difference metrics we analyze include CIE94 and CIE DE00 metrics based on the CIE L*a*b* color space. In addition\, we analyze d the newer ones derived for HDR applications: DEITP based on the ICTCP co lor space\, and DEz based on the Jzazbz color space. Since it’s generally agreed that color quality doesn’t change significantly when motion is adde d\, we used still image databases for evaluation.

\nTo quantif y the performance\, we use four standard performance evaluation procedures – Root mean square error\, Pearson linear correlation coefficient\, Spear man rank-order correlation coefficient and Outlier ratio. The color spaces derived for HDR were the best performers across the different databases\, but neither of those two metrics performed the best for every database. T hese databases have different experimental conditions and display specific ations. Analysis is currently underway to understand why the two different HDR color space metrics performed best in terms of these differing condit ions. END:VEVENT END:VCALENDAR