{"id":294,"date":"2022-04-22T13:27:15","date_gmt":"2022-04-22T03:27:15","guid":{"rendered":"http:\/\/www.computer-vision-software.com\/blog\/?p=294"},"modified":"2022-04-26T16:57:57","modified_gmt":"2022-04-26T06:57:57","slug":"pose-estimation-and-activity-recognition-demo","status":"publish","type":"post","link":"http:\/\/www.computer-vision-software.com\/blog\/2022\/04\/pose-estimation-and-activity-recognition-demo\/","title":{"rendered":"Pose Estimation and Activity recognition demo"},"content":{"rendered":"\n<figure class=\"wp-block-embed-youtube wp-block-embed is-type-video is-provider-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"wp-block-embed__wrapper\">\n<iframe loading=\"lazy\" title=\"Ambarella CV22 SoC-based Activity Recognition Demo\" width=\"500\" height=\"281\" src=\"https:\/\/www.youtube.com\/embed\/nVAGCLpMNGk?feature=oembed\" frameborder=\"0\" allow=\"accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share\" referrerpolicy=\"strict-origin-when-cross-origin\" allowfullscreen><\/iframe>\n<\/div><\/figure>\n\n\n\n<p>This demo showcases real-time Human Pose Estimation, based\non the Open Pose library, ported onto the camera platform, and designed by\nRhonda\u2019s Activity Recognition neural network for human behavior recognition.\nThe two Deep Learning Neural Networks (DNN),\nalong with the video pipeline, run on the Rhonda Software CV22 System on a Module\n(CV22 SoM).<\/p>\n\n\n\n<!--more-->\n\n\n\n<p>CV22 SoM &#8211; designed in-house\nas a low-power camera platform, is capable of running multiple neural\nnetworks, in addition to providing superior image quality. The core of the SoM platform is an Ambarella CV22 System\non a Chip \u2013 an ARM-based Image Signal Processor with a DNN inference\nacceleration engine, implemented on a single crystal.<\/p>\n\n\n\n<p>Both CV applications run simultaneously. The <em>Pose Estimation<\/em> network performs human body\ndetection in a full 4K frame, and people\u2019s figures recognized in the camera\u2019s field\nof view are visualized with \u201cskeleton-like\u201d pose markups. A blob of pixels\naround a foreground skeleton selected within the region of interest is passed\nto the <em>Activity recognition DNN<\/em>. <\/p>\n\n\n\n<p>The activity recognition algorithm is a simple, yet robust\ndemo built by Rhonda\u2019s CV team from scratch, and trained to identify several activity\ntypes: walking, standing, welcome hand gestures (high-five), jumping jacks,\nbody-weight squats. Recognized Activity for the foreground body is displayed in\nthe upper- left corner of the screen.<\/p>\n\n\n\n<p>After the initial port onto the CV22 platform, Open Pose\nalgorithm delivered a frame rate of 1 frame per second. It took a number of optimization\nprocedures performed by Rhonda\u2019s CV experts (such as pruning, quantization, and\ndedicated retraining) to achieve a fifteen fold acceleration in performance.<\/p>\n\n\n\n<p>The system can be trained for different use cases, such as\nsecurity, elderly care, production automation, sports activity analysis, and\nmore.<\/p>\n\n\n\n<p>For demo and testing purposes we&#8217;ve deployed a setup with HDMI video injection to show platforms&#8217; recognition capabilities with additional activities. <\/p>\n\n\n\n<figure class=\"wp-block-embed-youtube wp-block-embed is-type-video is-provider-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"wp-block-embed__wrapper\">\n<iframe loading=\"lazy\" title=\"Ambarella CV22 SoC-based Activity Recognition Demo with Injected Video Signal on the Camera Edge\" width=\"500\" height=\"281\" src=\"https:\/\/www.youtube.com\/embed\/Gd7Ymw_d85c?feature=oembed\" frameborder=\"0\" allow=\"accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share\" referrerpolicy=\"strict-origin-when-cross-origin\" allowfullscreen><\/iframe>\n<\/div><\/figure>\n\n\n\n<p>As a road-safety application example, Rhonda Software has assembled\nthe <em>Pedestrian detection<\/em> demo, based\non the same optimized port of the Open Pose library. The algorithm is applied\nto automotive conditions to detect pedestrians as participants of road traffic.<\/p>\n\n\n\n<figure class=\"wp-block-embed-youtube wp-block-embed is-type-video is-provider-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"wp-block-embed__wrapper\">\n<iframe loading=\"lazy\" title=\"Ambarella CV22 SoC-based Pedestrian Detection Demo\" width=\"500\" height=\"281\" src=\"https:\/\/www.youtube.com\/embed\/5C4vQ-MIT2E?feature=oembed\" frameborder=\"0\" allow=\"accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share\" referrerpolicy=\"strict-origin-when-cross-origin\" allowfullscreen><\/iframe>\n<\/div><\/figure>\n","protected":false},"excerpt":{"rendered":"<p>This demo showcases real-time Human Pose Estimation, based on the Open Pose library, ported onto the camera platform, and designed by Rhonda\u2019s Activity Recognition neural network for human behavior recognition. The two Deep Learning Neural Networks (DNN), along with the video pipeline, run on the Rhonda Software CV22 System on a Module (CV22 SoM).<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[56,39,38,84,35],"tags":[36,43,6],"class_list":["post-294","post","type-post","status-publish","format-standard","hentry","category-demo","category-demo-video","category-demo-videos","category-opencv","category-youtube","tag-object-recognition","tag-object-tracking","tag-opencv"],"_links":{"self":[{"href":"http:\/\/www.computer-vision-software.com\/blog\/wp-json\/wp\/v2\/posts\/294","targetHints":{"allow":["GET"]}}],"collection":[{"href":"http:\/\/www.computer-vision-software.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"http:\/\/www.computer-vision-software.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"http:\/\/www.computer-vision-software.com\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"http:\/\/www.computer-vision-software.com\/blog\/wp-json\/wp\/v2\/comments?post=294"}],"version-history":[{"count":0,"href":"http:\/\/www.computer-vision-software.com\/blog\/wp-json\/wp\/v2\/posts\/294\/revisions"}],"wp:attachment":[{"href":"http:\/\/www.computer-vision-software.com\/blog\/wp-json\/wp\/v2\/media?parent=294"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"http:\/\/www.computer-vision-software.com\/blog\/wp-json\/wp\/v2\/categories?post=294"},{"taxonomy":"post_tag","embeddable":true,"href":"http:\/\/www.computer-vision-software.com\/blog\/wp-json\/wp\/v2\/tags?post=294"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}