How can i use image data with sequential data for deep learning?

this is Bird-eye-view image.

I want to use this image with Target vehicle's histories that may be [x y velocity heading_angle] for forecasting vehicle's trajectory.

Is it possible to these data for deep learning? if u know how to do, please let me know.

thanks

Answer 1

[Shreeya]

LSTMs and GANs are usually employed for short/long term vehicle trajectory prediction. The bird's eye view data can help the model understand the future trajectory based on interaction with infrastructure or other vehicles.

The research paper linked down below details the usage of U-nets for an image to image regression for generating future visuals.

https://arxiv.org/abs/2207.01407