Get detection metrics

The get_metrics.py test script provide visual quantitive evaluations of both extraction methods. It use random, DQN and if available C51 agents, and computes the average and per class:

• Precision

• Recall

• F-score

• Intersection over Union (IOU)

It will run the game for some steps, and save the different metrics, as well as images for which the IOU threshold is met.

The parameters are:

• -g GAME, --game GAME : game to evaluate (e.g. ‘Pong’)

• -i IOU, --iou IOU : Minimum iou threshold to trigger image saving (e.g. 0.7)

• -p PATH, --path PATH : the path to a potential DQN agents’ model

• -s SEED, --seed SEED If provided, set the seed