Artificial intelligence (AI) is a technology that is becoming more and more prevalent in online tools today. And for good reason, it allows you to perform certain tasks in record time. It therefore greatly facilitates the life of human beings.
In this sense, I wanted to introduce you today Which Frame , a new AI-based online tool that allows you to quickly find an image within a video using a simple semantic search, such as those that we do in Google.
For example, if you want to find a scene showing a blue car in a video, all you have to do is type in the keyword “blue car” and the Which Frame AI will find all the images related to your search.
The Which Frame site was created by a certain David Chuan-en Lin, a student at Carnegie Mellon University based in Pittsburgh in the United States. Its web application is powered by OpenAI’s CLIP neural network and the interface was built with Streamlit .
To use the service, no registration is required. All you have to do is go to Which Frame’s site and upload your video in MP4 format with a maximum size of 20 MB or indicate the URL of a YouTube video of up to 5 minutes. The system will then analyze your video for a few seconds.
Then choose the search method you want: Text , Image or Text + Image . If you choose the first method, type the keyword you want just below, then click on the ” Submit ” button to start the search.
After a few seconds, the service will show you all the available results with, of course, the possibility of downloading the images in question. So you can find specific frames in any video with minimal effort.