{"id":13219,"date":"2020-08-13T14:20:15","date_gmt":"2020-08-13T21:20:15","guid":{"rendered":"http:\/\/jnack.com\/blog\/?p=13219"},"modified":"2025-03-12T18:59:04","modified_gmt":"2025-03-13T01:59:04","slug":"body-movin-google-ai-releases-new-tech-for-body-tracking-eye-measurement","status":"publish","type":"post","link":"http:\/\/jnack.com\/blog\/2020\/08\/13\/body-movin-google-ai-releases-new-tech-for-body-tracking-eye-measurement\/","title":{"rendered":"Body Movin’: Google AI releases new tech for body tracking, eye measurement"},"content":{"rendered":"\n

My old teammates keep slapping out the bangers, releasing machine-learning tech to help build apps that key off the human form.<\/p>\n\n\n\n

First up is Media Pipe Iris<\/a>, enabling depth estimation for faces without fancy (iPhone X-\/Pixel 4-style) hardware, and that in turn opens up access to accurate virtual try-on for glasses, hats, etc.:<\/p>\n\n\n\n

\nhttps:\/\/twitter.com\/GoogleAI\/status\/1291430839088103424\n<\/div><\/figure>\n\n\n\n

The model enables cool tricks like realtime eye recoloring:<\/p>\n\n\n\n

\"\"\/<\/figure>\n\n\n\n

I always find it interesting to glimpse the work that goes in behind the scenes. For example:<\/p>\n\n\n\n

To train the model from the cropped eye region, we manually annotated ~50k images, representing a variety of illumination conditions and head poses from geographically diverse regions, as shown below.<\/p><\/blockquote>\n\n\n\n

\"\"\/<\/figure>\n\n\n\n

The team has followed up this release with MediaPipe BlazePose<\/a>, which is in testing now & planned for release via the cross-platform ML Kit soon:<\/p>\n\n\n

\"\"\"\"<\/p>\n\n\n

Our approach provides human pose tracking by employing machine learning (ML) to infer 33, 2D landmarks of a body from a single frame. In contrast to current pose models based on the standard COCO topology<\/a>, BlazePose accurately localizes more keypoints, making it uniquely suited for fitness applications…<\/p>

If one leverages GPU inference, BlazePose achieves super-real-time performance, enabling it to run subsequent ML models, like face or hand tracking.<\/p><\/blockquote>\n\n\n\n

Now I can’t wait for apps to help my long-suffering CrossFit coaches actually quantify<\/em> the crappiness of my form. Thanks, team! 😛 <\/p>\n","protected":false},"excerpt":{"rendered":"

My old teammates keep slapping out the bangers, releasing machine-learning tech to help build apps that key off the human form. First up is Media Pipe Iris, enabling depth estimation for faces without fancy (iPhone X-\/Pixel 4-style) hardware, and that in turn opens up access to accurate virtual try-on for glasses, hats, etc.: The model […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":[],"categories":[13,17,77],"tags":[],"_links":{"self":[{"href":"http:\/\/jnack.com\/blog\/wp-json\/wp\/v2\/posts\/13219"}],"collection":[{"href":"http:\/\/jnack.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"http:\/\/jnack.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"http:\/\/jnack.com\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"http:\/\/jnack.com\/blog\/wp-json\/wp\/v2\/comments?post=13219"}],"version-history":[{"count":8,"href":"http:\/\/jnack.com\/blog\/wp-json\/wp\/v2\/posts\/13219\/revisions"}],"predecessor-version":[{"id":13258,"href":"http:\/\/jnack.com\/blog\/wp-json\/wp\/v2\/posts\/13219\/revisions\/13258"}],"wp:attachment":[{"href":"http:\/\/jnack.com\/blog\/wp-json\/wp\/v2\/media?parent=13219"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"http:\/\/jnack.com\/blog\/wp-json\/wp\/v2\/categories?post=13219"},{"taxonomy":"post_tag","embeddable":true,"href":"http:\/\/jnack.com\/blog\/wp-json\/wp\/v2\/tags?post=13219"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}