We are exploring ways to use video compression more effectively, by focusing on places in space and time that are more important and making sure they are compressed better than other less important places.
We would like to build a video annotation tool that can run from a browser and allow human annotators to specify which areas are more important. Specifically, given a video of dimensions H x W and T frames, we want to create an "importance map". It is an array of size H x W x T and values 0..1 which tells us how important each pixel is. As we annotate the importance, we will use it to compress the video interactively.
This is an open-ended project. We will be developing the UI as we go along and explore what works. We are looking for someone who can create quick and high quality prototypes and who can work on this ideally 8 hours a day, so we can make quick progress.
The first step is to create a UI that has two windows. On the left we display a 2-second video playing in a loop. On the right we display one frame of the video, and have a slider that we can drag to move back and forth.
Please estimate the amount of hours you will need to prototype this.
18 freelances font une offre moyenne de 34 $/heure pour ce travail
Hello there. I checked your proposal. I worked in similar project like yours. [login to view URL] Please feel free to contact me so we can start right now. Regards. Jered.