{"id":7291,"date":"2022-08-10T22:00:06","date_gmt":"2022-08-11T04:00:06","guid":{"rendered":"https:\/\/roc.ai\/?p=7291"},"modified":"2022-08-10T22:00:06","modified_gmt":"2022-08-11T04:00:06","slug":"rank-one-computing-ranks-1-in-latest-nist-frvt-for-combined-accuracy-and-efficiency","status":"publish","type":"post","link":"https:\/\/roc.ai\/2022\/08\/10\/rank-one-computing-ranks-1-in-latest-nist-frvt-for-combined-accuracy-and-efficiency\/","title":{"rendered":"Rank One Computing Ranks #1 in Latest NIST FRVT for Combined Accuracy and Efficiency"},"content":{"rendered":"

[et_pb_section fb_built=”1″ _builder_version=”4.16″ collapsed=”off” global_colors_info=”{}”][et_pb_row admin_label=”Opening Section” _builder_version=”4.16″ _module_preset=”default” custom_padding=”0px||||false|false” locked=”off” collapsed=”off” global_colors_info=”{}”][et_pb_column type=”4_4″ _builder_version=”4.16″ _module_preset=”default” global_colors_info=”{}”][et_pb_text admin_label=”Opening Section” _builder_version=”4.21.0″ _module_preset=”2e74e5d0-f316-42aa-a348-8f9045cb3398″ text_text_color=”#090E18″ hover_enabled=”0″ global_colors_info=”{}” content__hover_enabled=”off|desktop” sticky_enabled=”0″]<\/p>\n

ROC\u2019s SDK v2.2 is now the single-highest performing vendor of all U.S. FR algorithm providers in the latest National Institute of Standards and Technology (NIST) Face Recognition Vendor Test (FRVT). Specifically, v2.2 outranks all others with both top tier accuracy and efficiency in both the 1:1 and 1:N series. At 2-5x faster and 10-20x more efficient, v2.2 packs a punch of impressive computer vision developments: 5x improvement in liveness fraud detection, expanded PersonID searching and brand new LPR functionality. These performance improvements will directly enhance performance of ROC Watch and all ROC applications. Contact us to ensure you have up to date versions of our top-ranked SDK<\/a>.<\/span><\/p>\n

True to our company culture, ROC continues to deliver the most efficient, accurate, and trusted computer vision algorithms available on the market today.\u00a0 I\u2019m incredibly proud of the work we have done in a few short months since releasing our next generation AI\/ML-powered version 2.0. In that release, we included some exciting new computer vision capabilities including <\/span>Liveness<\/b> fraud prevention, <\/span>Vehicle<\/b> and <\/span>Object Detection<\/b> and <\/span>License Plate Recognition (LPR)<\/b> based on <\/span>Optical Character Recognition (OCR)<\/b>. Read more about v2.0 in our<\/span> previous blog<\/span><\/a>.<\/span><\/p>\n

With the v2.2 release, our AI\/ML developers have made impressive gains that deliver powerful improvements to <\/span>Liveness <\/b>fraud prevention, <\/span>LPR<\/b>, and <\/span>Person ID <\/b>in addition to ground-breaking accuracy performance for Face Recognition<\/a> – all in a single, efficient SDK packet. <\/span><\/p>\n

[\/et_pb_text][\/et_pb_column][\/et_pb_row][et_pb_row admin_label=” H2-Subheader” _builder_version=”4.16″ _module_preset=”default” custom_padding=”15px||0px||false|false” locked=”off” collapsed=”off” global_colors_info=”{}”][et_pb_column type=”4_4″ admin_label=”Column” _builder_version=”4.16″ _module_preset=”default” global_colors_info=”{}”][et_pb_text admin_label=”FR Improvements” _builder_version=”4.16″ _module_preset=”b713cdde-22bd-4fec-aa50-5b35c5128db3″ global_colors_info=”{}”]<\/p>\n

Face Recognition\u00a0 Improvements<\/b><\/p>\n

[\/et_pb_text][\/et_pb_column][\/et_pb_row][et_pb_row column_structure=”3_5,2_5″ admin_label=”H2-Text” _builder_version=”4.16″ _module_preset=”default” custom_padding=”15px||0px||false|false” locked=”off” collapsed=”off” global_colors_info=”{}”][et_pb_column type=”3_5″ admin_label=”Column” _builder_version=”4.16″ _module_preset=”default” global_colors_info=”{}”][et_pb_text admin_label=”FR Improvements” _builder_version=”4.16″ _module_preset=”default” global_colors_info=”{}”]<\/p>\n

While accuracy metrics are the primary currency of FR vendors, integrators of FR systems know that real-world performance of these algorithms requires that they are deployable in real-world settings, such as high throughput video processing, large-scale enterprise search systems, or on-edge mobile use-cases. For this reason, ROC continues to prioritize speed and efficiency in parallel to our continued drive for top accuracy performance.\u00a0 And with v2.2, we have continued to establish ourselves as the industry leader. <\/span><\/p>\n

[\/et_pb_text][\/et_pb_column][et_pb_column type=”2_5″ admin_label=”Column” _builder_version=”4.16″ _module_preset=”default” global_colors_info=”{}”][et_pb_text admin_label=”Sidebar” _builder_version=”4.16″ _module_preset=”default” text_text_color=”#FFFFFF” background_color=”#FD4D2D” custom_padding=”5px|10px|15px|10px|false|false” global_colors_info=”{}”]<\/p>\n

Just as smart phones require chips that are small, efficient, and fast – FR applications require algorithms that are efficient, fast, and accurate.\u00a0 With ROC SDK v2.2, you don\u2019t have to choose between accuracy and efficiency.<\/i><\/b><\/span><\/p>\n

[\/et_pb_text][\/et_pb_column][\/et_pb_row][et_pb_row admin_label=”H3 – FR Accuracy Charts” _builder_version=”4.16″ _module_preset=”default” custom_padding=”10px||0px||false|false” locked=”off” collapsed=”on” global_colors_info=”{}”][et_pb_column type=”4_4″ _builder_version=”4.16″ _module_preset=”default” global_colors_info=”{}”][et_pb_text _builder_version=”4.16″ _module_preset=”default” global_colors_info=”{}”]<\/p>\n

As shown in Figure 1 below from the 07\/28\/2022 NIST FRVT Ongoing report, Rank One is, by a wide margin, able to deliver the best combination of accuracy and efficiency when measuring the mean FRVT rankings in all eight accuracy benchmarks (Tables <\/span>18 to 27: <\/span>Visa MC, Visa, Mugshot, Mugshot 12+Yrs, VisaBorder, Border 10-6, Border 10-5, Wild) and all four efficiency benchmarks (Tables 8 to 17: template generation speed, template size, binary size, and comparison speed.)<\/span><\/p>\n

[\/et_pb_text][et_pb_text admin_label=”Title 1″ _builder_version=”4.16″ _module_preset=”default” header_text_align=”center” header_text_color=”#1F1F56″ custom_margin=”||||false|false” global_colors_info=”{}”]<\/p>\n

Figure 1: Mean Accuracy and Efficiency Ranking across 28 Commercial FR Vendors<\/span><\/h5>\n

[\/et_pb_text][et_pb_image src=”https:\/\/roc.ai\/wp-content\/uploads\/2022\/08\/FRVT-blog-Figure-1.png” title_text=”FRVT blog Figure 1″ force_fullwidth=”on” admin_label=”Chart 1″ _builder_version=”4.16″ _module_preset=”default” global_colors_info=”{}”][\/et_pb_image][et_pb_text admin_label=”Text” _builder_version=”4.16″ _module_preset=”default” global_colors_info=”{}”]<\/p>\n

When focusing specifically on accuracy metrics, v2.2 continues ROC\u2019s impressive climb in rankings.\u00a0 For example, as shown in Figure 2 below, ROC SDKv2.2 is ranked as the #1 most accurate U.S. vendor on Border crossing data <\/strong>(1E-5). v2.2 also ranked #6 globally of 440 algorithms total with all of the top 5 developed in China, Russia, and Korea.<\/p>\n

[\/et_pb_text][et_pb_text admin_label=”Title 2″ _builder_version=”4.16″ _module_preset=”default” header_text_align=”center” header_text_color=”#1F1F56″ custom_margin=”||||false|false” global_colors_info=”{}”]<\/p>\n

Figure 2: Error Rates for on Border Data (1E-5) for U.S. Vendors<\/span><\/h5>\n

[\/et_pb_text][et_pb_image src=”https:\/\/roc.ai\/wp-content\/uploads\/2022\/08\/FRVT-blog-Figure-2-1.png” title_text=”FRVT blog Figure 2″ show_bottom_space=”off” force_fullwidth=”on” admin_label=”Chart 2″ _builder_version=”4.16″ _module_preset=”default” custom_margin=”||||false|false” global_colors_info=”{}”][\/et_pb_image][et_pb_text _builder_version=”4.16″ _module_preset=”default” global_colors_info=”{}”]<\/p>\n

And more broadly, ROC v2.2 achieved better than 99.5% genuine match rate on all Border, Visa, and Mugshot datasets. Figure 3 demonstrates this as well as the dramatic 1.4x improvements that ROC has made in just the last six months.\u00a0<\/p>\n

[\/et_pb_text][et_pb_text admin_label=”Title 3″ _builder_version=”4.16″ _module_preset=”default” header_text_align=”center” header_text_color=”#1F1F56″ custom_margin=”||||false|false” global_colors_info=”{}”]<\/p>\n

Figure 3: ROC SDK v2.2 Error rate improvements over v2.0<\/span><\/h5>\n

[\/et_pb_text][et_pb_image src=”https:\/\/roc.ai\/wp-content\/uploads\/2022\/08\/FRVT-blog-Figure-3-1.png” title_text=”FRVT blog Figure 3″ force_fullwidth=”on” admin_label=”Chart 3″ _builder_version=”4.16″ _module_preset=”default” custom_margin=”||||false|false” global_colors_info=”{}”][\/et_pb_image][\/et_pb_column][\/et_pb_row][et_pb_row admin_label=”H4 – Liveness” module_id=”identity-verification” _builder_version=”4.16″ _module_preset=”default” custom_padding=”20px||0px||false|false” locked=”off” collapsed=”on” global_colors_info=”{}”][et_pb_column type=”4_4″ admin_label=”Column” _builder_version=”4.16″ _module_preset=”default” global_colors_info=”{}”][et_pb_text admin_label=”Liveness Header” _builder_version=”4.16″ _module_preset=”cbd6671a-637c-4930-92b6-e8cc746839c2″ global_colors_info=”{}”]<\/p>\n

Liveness Fraud Prevention<\/strong><\/p>\n

[\/et_pb_text][et_pb_text admin_label=”Liveness Body” _builder_version=”4.16″ _module_preset=”2e74e5d0-f316-42aa-a348-8f9045cb3398″ text_text_color=”#090E18″ global_colors_info=”{}”]<\/p>\n

As defined by the Presentation Attack Detection (PAD) standards outlined in <\/span>ISO\/IEC 30107-3<\/span><\/a>, Liveness detection seeks to combat the following attempts to make a fraudulent FR match:<\/span><\/p>\n