To complete the project we need to do the following task one by one.
-
Create frames of given video, first one every 5 seconds.
-
Detect faces from the frames. Should be able to detect multiple faces in one image.
-
Detect hyper emotion from the faces. (In progress. Prototype can be expected in 2-3 days to a maximum of 4-5 days.)
-
Use timestamps or array or something to store the exact time/frame for the trim.
-
For each hyper emotion(more like so much emotion), from the original video create a mini video containing 2 secs before the emotion and 2 secs after.
-
Include audio in those mini videos. Either don't remove it during the trim or add it appropriately as per the original later.
-
Combine all the mini videos into a single.
-
Perform Test Driven Development (TDD) using any testing framework.
-
Create documentation for the project.
- Use sphinx autodocs.
- Write function docstrings in numpy format.