Shulou(Shulou.com)06/02 Report--

Previously introduced a lot of Microsoft Azure AI and machine learning capabilities for Demo for demonstration, this time introduce a website that can be used for powerful video analysis: AMS video indexer

On this website, you can upload a local video for video analysis. After uploading successfully, you can do a lot of analysis in different dimensions according to this video. For example, here I upload a video from my local computer and after a while of analysis and retrieval, I can click and play it.

After playing, you will see and analyze how many people have appeared in the video, how many seconds or an interval each person has appeared in the whole video, and how many topics appear in the video.

And identify which keywords appear frequently in the video; all the object tags that have appeared, such as people, men, outside, windows, walls, etc.; also identify what brands appear in the video, and all these tags or brands are marked in the number of seconds or intervals in the video, which can be quickly clicked to switch to these seconds.

Also identify the feelings of a certain person in the whole video, how much they are happy, and so on.

It also identifies how many scenes appear in the video and the keyframes of the character I selected appear throughout the video, which can be clicked to fast forward and rewind to the frame or second in which the person appears.

Secondly, the subtitle of the whole video can be realized by voice-to-text in the timeline, and the subtitle scrolls automatically with the voice played.

And the text can also be translated into the language that needs to be displayed.

Of course, you can also check the location of various tags interspersed in the entire event line, and edit the content of the text for content revision.

Finally, you can download scripts, insights, and translated txt texts.

If you download the VTT file, you can play the unanalyzed source video locally with subtitles (choose the downloaded VTT file) through the local player. Does it feel very simple?

From this video indexing analysis, the back end is actually combined with a variety of Microsoft Azure AI+ machine learning services for comprehensive use, such as the use of computer vision to compare scenes and objects, face recognition video faces, voice services, English subtitles for videos without subtitles, and translation services to translate English into the languages you need, such as Chinese, etc. Emotional recognition is used to classify the percentage of the characters in the video, so it is very interesting to analyze different opinions from multiple dimensions from a simple video. If you are interested, you can also do it yourself.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.