Question regarding multi-task learning in computer vision

For instance, if I am using multi-task learning to train a model for a self driving car, how can I know which part of the image got detected or in other words how to localize the detected object (or objects) in that image?

The idea is to use boxes and also predicting their parameters to localize the objects. You’ll see that more in detail in course 4 :slight_smile:

1 Like