Classification is performed on each individual seat. Consequently, the images need to be split into three rectangles such that each seat can be classified individually (see left and right image). For cars with only two seats, the middle one is not used (see middle image).
One could decide to classify a seat with an everyday object (and an empty infant/child seat) as empty as well. Further, one can train a single classifier for the three seats, or, for example, train one classifier for each seat.
The following table reports the public leaderboard for different training data and vehicle used. We use the following abbreviations for the classes:
- BG = background
- IS = infant seat
- CS = child seat
- Person = Adult passenger
- Object = everyday object
- E-IS = empty infant seat
- E-CS = empty child seat
Train car all means that one model was trained on each vehicle. The general performance of the method is evaluated on the test set of each vehicle without the test performance of the vehicle it was trained on. Consequently, we calculate the mean of the means of the performances across all vehicles for the overall performance of the method.
If a single car is mentioned as the car the model was trained on, then a single model was trained only on the mentioned car and the performance of this model on the test images of all unseen/unknown vehicles is evaluated. Consequently, we calculate the mean of the means of the performances across all vehicles without the test performance of the vehicle it was trained on.