Week 6. Buffer + Gap

In order to keep the information about the input values we need to create a buffer that changes every new frame. We need this buffer to predict the input + gap(30) frame and to allways show the route the object is taking in the GUI

This week I have been improving how we extract the data in the best way posible, for that we have added a new parameter to the model, the number of the frame. The problem of prediccition should be solved taking more detailed data from the videos in the following way:

Nº Frame 0 - Buffer Array [0-19] - Predict frame 50 Nº Frame 1 - Buffer Array [1-20] - Predict frame 51 Nº Frame 2 - Buffer Array [2-21] - Predict frame 52 Nº Frame 3 - Buffer Array [3-22] - Predict frame 53 Nº Frame 4 - Buffer Array [4-23] - Predict frame 54 ……

Next steps are to resolve the inference of the Network by creating a similar dataset and trainging a model that matches the data extracted from the videos. We need to add the new parameter nº of frame