Face recognition - learning from manual input not accurate

i used video indexer for a video with a celebrity in it. It identified the celebrity correctly in some parts of the video. But marked him as unknown person in other parts(4 seperate unknown people). I manually tagged the other unknowns also by the same celebrity name. When I run it again, it correctly says that the celebrity is in 91% of the video (combines all tags) but still shows the other 4 as the celebrity, but found in 0% of the video.



    I'm from the VI team, thanks for reaching out! We will look into your case and update when we have insights.
    Posted by Hidden Mon, 04 Feb 2019 08:38:43 GMT
    Hi, this is a client issue. It will be fixed in the next couple weeks. If you export the JSON you will see the right %.
    Posted by Hidden Mon, 04 Feb 2019 13:27:13 GMT

You're not signed in. Please sign-in to report an issue or post a comment.