We are given a data synthesizer that generates images and labels. The goal is to train a model with at most 4.5 million trainable parameters which determines whether each image has a star and, if so, finds a rotated bounding box that bounds the star (as shown below).
More precisely, the labels contain the following five numbers, which the model should predict:
- the x and y coordinates of the center
- yaw
- width and height.
If there is no star, the label consists of 5 np.nans. The height of the star is always noticeably larger than its width, and the yaw points in one of the height directions. The yaw is always in the interval [0, 2 * pi), oriented counter-clockwise and with zero corresponding to the upward direction. train.py contains a basic CNN architecture (and training code) that performs fairly and you can extend this model/training or start over on your own.
