The design is solely evaluated by the total energy consumption; however, there are minimum thresholds on accuracy and throughput. If these minimums are not met, penalties are applied.
Accuracy
Accuracy is measured using IoU (Intersection over Union). A good example of IoU can be found at https://www.pyimagesearch.com/2016/11/07/intersection-over-union-iou-for-object-detection/.
The minimum accuracy should be 0.7, otherwise a penalty is applied.
Throughput.
The design should achieve at least 30 FPS, otherwise a penatly is applied.
Scoring Function
The score for a team is calculated as follows:
Score = 10^2 / log2(Energy) × Max(ReLU([1 - 5 × ReLU(0.7 - IoU)]), 0.1) × ReLU([1 - ReLU(1 - FPS / 30)])
ReLU is a non-linear function that helps apply the penalty: ReLU(x) = (x > 0) ? x : 0.