Today at the social event we finally got our results for the functionality benchmarks.
In object perception we got 3rd place. In speech recognition we got 3rd place. In navigation we got 1st place.
During the day the task benchmarks were running.
The first challenge,getting to know your home, was about recognizing changes in the test enviroment and reporting them with help of as semantic map. In the first try we had a bug in writing the changes to a file - therefore we didn’t get much points here. We fixed the bug before the second try and therefore were able to score 3 achievements and got 0 penalties.
The second challenge, welcoming visitors, tested recognition and case handling of persons ringing the doorbell. In the first try we detected faces in the background and got wrong recognitions. But on the second try we were able to recognize all but one person right and handle their cases. Therefore we scored 6 points and 0 penalties.
The last challenge, catering for granny annie’s comfort, is near to a general service robot. So the owner - in this case granny - gives commands to the robot. The robot tries to execute every command it gets. Lisa got into a state loop in the first try and wasn’t able to execute every command. But in the second try we performed better. We had some problems understanding granny and managed to hit a plant. But luckily we were allowed to continue and show what Lisa can do. So we got 3 achievements and 2 penalties.
With this results we are proceeding into round three tomorrow - so we will use every time we can get to finish the latest hacks for the final rounds of the competition tomorrow.
cite of the day “home sweet home arena” from unknown bystander