combing1のブログ -2ページ目

combing1のブログ

ブログの説明を入力します。

Bole Online [REVIEW]: Recently Microsoft How-old.net Air Max 2011 Womens Purple Black Grey website very hot, Nike Free Run 3 Women users upload photos after the Air Jordan Outlet station, it can measure out the photo figures gender and age. Some foreign friends on Quora asked How-old.net works. Below are involved Eason Wang Microsoft project manager of the project reply. (Eason 579756 403 Purple White Nike Black Mamba 24 Kobe Sale Wang is a senior program manager of Bing, this answer received three thousand three hundred point Like) I am directly involved in the project. To be honest, this popular small site far beyond my expectations. I have done some analysis afterwards why popular and Medium wrote an article. Back to the topic, my answer into two parts. The first part will talk about how to quickly achieve exactly the same in any of the software features, the second part I will go into some of the description of the technology itself. In the past few years and the cooperation of Microsoft's R \u0026 D department, on Bing's image search technology is concerned, we have reached the best industrial image understanding, rapid extension of this technology to other Microsoft products. Microsoft is currently in the Oxford project home, this technology is open to all developers. Want to achieve the same functionality in their software, you Air Max 2011 Womens Purple Grey simply need to call at our web API, you can get all the information needed in JSON format. You can try this page in www.projectoxford.ai upload Air Max 2012 Black Navy Blue White an image, it will give you results within a few seconds, face coordinates, gender and age information are inside. Face API is only one we did in the Oxford project characteristics. There are many other core features to help build innovative applications. Microsoft's internal API open to the public so I am very excited, I knew this would develop community play a profound impact. Seemingly impossible things before it makes only a simple web API call can be done about. #HowOldRobot Just a small showcase these capabilities, Azure machine learning team a developer only one day put it developed. (Translator's Note: The following is the API Samples for JSON format) JSON: [{\u0026 quot; faceId \u0026 quot ;: \u0026 quot; 5af35e84-ec20-4897-9795-8b3d4512a1f9 \u0026 quot ;, \u0026 quot; faceRectangle \u0026 quot ;: {\u0026 quot; width \u0026 quot ;: 60, \u0026 quot ; height \u0026 quot ;: 60, \u0026 quot; left \u0026 quot ;: 276, \u0026 quot; top \u0026 quot ;: 43}, \u0026 quot; faceLandmarks \u0026 quot ;: {\u0026 quot; pupilLeft \u0026 quot ;: {\u0026 quot; x \u0026 quot ;: \u0026 quot; 295.1 \u0026 quot ;, \u0026 quot; y \u0026 quot ;: \u0026 quot ; 56.8 \u0026 quot;}, \u0026 quot; pupilRight \u0026 quot ;: {\u0026 quot; New Nike Free Run 3 Shoes Silver 3 x \u0026 quot ;: \u0026 quot; 317.9 \u0026 quot ;, \u0026 quot; y \u0026 quot ;: \u0026 quot; 59.6 \u0026 quot;}, \u0026 quot; noseTip \u0026 quot ;: {\u0026 quot; x \u0026 quot ;: \u0026 quot; 579756 403 Purple White Nike Black Mamba 24 Kobe Sale 311.6 \u0026 quot ;, \u0026 quot; y \u0026 quot ;: \u0026 quot; 74.7 \u0026 quot;}, \u0026 quot; mouthLeft \u0026 quot ;: {\u0026 quot; x \u0026 quot ;: \u0026 597806-400 Nike LeBron X EXT QS Denim-Pink Outlet quot; 291.0 \u0026 quot ;, \u0026 quot; y \u0026 quot ;: \u0026 quot; 86.3 \u0026 quot;}, \u0026 quot; mouthRight \u0026 quot ;: {\u0026 quot ; x \u0026 quot ;: \u0026 quot; 311.6 \u0026 quot ;, \u0026 quot; y \u0026 quot ;: \u0026 quot; 88.6 \u0026 quot;}, \u0026 quot; eyebrowLeftOuter \u0026 quot ;: {\u0026 quot; x \u0026 quot ;: \u0026 quot; 281.6 \u0026 quot ;, \u0026 quot; y \u0026 quot ;: \u0026 quot; 50.1 \u0026 quot;} , \u0026 quot; eyebrowLeftInner \u0026 quot ;: {\u0026 quot; x \u0026 quot ;: \u0026 quot; 304.2 \u0026 quot ;, \u0026 quot; y \u0026 quot ;: \u0026 quot; 51.6 \u0026 quot;}, \u0026 quot; eyeLeftOuter \u0026 quot ;: {\u0026 quot; x \u0026 quot ;: \u0026 quot; 289.1 \u0026 quot ;, \u0026 quot; y \u0026 quot ;: \u0026 quot; 57.1 \u0026 quot;}, Air Max 2011 Womens Purple Black Grey \u0026 quot; eyeLeftTop \u0026 quot ;: {\u0026 quot; x \u0026 quot ;: \u0026 quot; 294.0 \u0026 quot ;, \u0026 quot; y \u0026 quot ;: \u0026 quot; 54.5 \u0026 quot;}, \u0026 quot; eyeLeftBottom \u0026 quot ;: {\u0026 quot; x \u0026 quot ;: \u0026 quot ; 293.0 \u0026 quot ;, \u0026 quot; y \u0026 quot ;: \u0026 quot; 61.0 \u0026 quot;}, \u0026 quot; eyeLeftInner \u0026 quot ;: {\u0026 quot; x \u0026 quot ;: \u0026 quot; 297.8 \u0026 quot ;, \u0026 quot; y \u0026 quot ;: \u0026 quot; 58.7 \u0026 quot;}, \u0026 quot; eyebrowRightInner \u0026 quot; : {\u0026 quot; x \u0026 quot ;: \u0026 quot; 316.0 \u0026 quot ;, \u0026 quot; y \u0026 quot ;: \u0026 quot; 54.2 \u0026 quot;}, \u0026 quot; eyebrowRightOuter \u0026 quot ;: {\u0026 quot; x \u0026 quot ;: \u0026 quot; 324.7 \u0026 quot ;, \u0026 quot; y \u0026 quot ;: \u0026 quot; 54.2 \u0026 quot;}, \u0026 quot; eyeRightInner \u0026 quot ;: {\u0026 quot; x \u0026 quot ;: \u0026 quot; 312.9 \u0026 quot ;, \u0026 quot; y \u0026 quot ;: \u0026 quot; 60.9 \u0026 quot;}, \u0026 quot; eyeRightTop \u0026 quot 586590-300 Nike Kobe 8 System Elite GC Poison Green Superhero Outlet ;: {\u0026 quot; x \u0026 New Nike Free 5.0 V4 Purple Yellow Shoes quot ;: \u0026 quot; 317.8 \u0026 quot ;, \u0026 quot; y \u0026 quot ;: \u0026 quot; 57.7 \u0026 quot;}, \u0026 quot; eyeRightBottom \u0026 quot ;: {\u0026 quot; x \u0026 quot ;: \u0026 quot; 317.9 \u0026 quot ;, \u0026 quot; y \u0026 quot ;: \u0026 quot; 63.7 \u0026 quot;}, \u0026 quot; eyeRightOuter \u0026 quot ;: {\u0026 quot; x \u0026 quot ;: \u0026 quot; 322.8 \u0026 quot ;, \u0026 quot; y \u0026 quot ;: \u0026 quot; 60.8 \u0026 quot;}, \u0026 quot; noseRootLeft \u0026 quot ;: {\u0026 quot; x \u0026 quot Air Max 2011 Womens Grey Green ;: \u0026 quot; 304.0 \u0026 quot ;, \u0026 quot; y \u0026 quot ;: \u0026 quot; 60.2 \u0026 quot;}, \u0026 quot ; noseRootRight \u0026 quot ;: {\u0026 quot; x \u0026 quot ;: \u0026 quot; 312.2 \u0026 quot ;, \u0026 quot; y \u0026 quot ;: \u0026 quot; 61.2 \u0026 quot;}, \u0026 quot; noseLeftAlarTop \u0026 quot ;: {\u0026 quot; x \u0026 quot ;: \u0026 quot; 302.6 \u0026 quot ;, \u0026 quot; y \u0026 quot ;: \u0026 quot; 70.2 \u0026 quot;}, \u0026 quot; noseRightAlarTop \u0026 quot ;: {\u0026 quot; x \u0026 quot ;: \u0026 quot; 313.0 \u0026 quot ;, \u0026 quot; y \u0026 quot ;: \u0026 quot; 70.0 \u0026 quot;}, \u0026 quot; noseLeftAlarOutTip \u0026 quot ;: {\u0026 quot; x \u0026 quot ;: \u0026 quot; 298.8 \u0026 quot ;, \u0026 quot; y \u0026 quot ;: \u0026 quot; 76.2 \u0026 quot;}, \u0026 Nike Air Max quot; noseRightAlarOutTip \u0026 quot ;: {\u0026 quot; x \u0026 quot ;: \u0026 quot; 315.2 \u0026 quot ;, \u0026 quot; y \u0026 quot ;: \u0026 quot; 76.6 \u0026 quot;}, \u0026 quot; upperLipTop \u0026 quot ;: { \u0026 quot; x \u0026 quot ;: \u0026 quot; 307.3 \u0026 quot ;, \u0026 quot; y \u0026 quot ;: \u0026 quot; 84.0 \u0026 quot;}, \u0026 quot; upperLipBottom \u0026 quot ;: {\u0026 quot; x \u0026 quot ;: \u0026 quot; 306.6 \u0026 quot ;, \u0026 quot; y \u0026 quot ;: \u0026 quot; 86.4 \u0026 quot; }, \u0026 quot; underLipTop \u0026 quot ;: {\u0026 quot; x \u0026 quot ;: \u0026 quot; 305.5 \u0026 quot ;, 2015 Nike Free 5.0 \u0026 quot; y \u0026 quot ;: \u0026 quot; 89.6 \u0026 quot;}, \u0026 quot; underLipBottom \u0026 quot ;: {\u0026 quot; x \u0026 quot ;: \u0026 quot; 304.1 \u0026 quot ;, \u0026 quot; y \u0026 quot ;: \u0026 quot; 94.0 \u0026 quot;}}, \u0026 quot; attributes \u0026 quot ;: {\u0026 quot; age \u0026 quot ;: 24, \u0026 quot; gender \u0026 quot ;: \u0026 quot; female \u0026 quot ;, \u0026 quot; headPose \u0026 quot ;: {\u0026 quot; roll \u0026 quot ;: \u0026 quot; 4.0 \u0026 quot; , \u0026 quot; yaw \u0026 quot ;: \u0026 quot; 31.3 \u0026 quot ;, \u0026 quot; pitch \u0026 quot ;: \u0026 quot; 0.0 \u0026 quot;}}}] How-old.net this site mainly depends on three key technologies: face detection, sex and age detection ʱ?? Face detection is the basis for the other two. Age and sex tests for the detection, the only machine learning typical regression and classification problems, related to the representation of facial features, build collect training data, regression and classification model and optimization model. There are many in this area have been published in the papers. If you are interested to know more about you tell me. On the other hand, deep learning and understanding of large-scale data driven image understanding of breakthrough for smarter systems and procedures port opens the door. You can look at my latest chart on how to apply more advanced image scenes blog: http: //blogs.bing.com/search-qua... Leave your comment let me improve this answer, thank you!How-old.net website How does it work?