Overcoming 17 teams at the school round “Digital Race 2018 – 2019” held at FPT University in Hanoi, C213 and Rolling Thunder are the two names that will represent FPT University to participate in the semifinal with the theme of “Autonomous Car”, competing with students of colleges and universities across the country.
Here, the teams have experienced 3 parts: Examining the ability to process images; Debate; Quick programming to find teams that excel in creativity and other skills. In particular, the number of points for each section accounted for 40%, 30%, and 30% respectively. After finishing 3 sections, 2 teams with the highest total scores will be the winner and represent the school entering the semi-final round.
At the round testing their image processing ability, organizers have designed a simulation software and within 30 minutes, the teams will use their algorithms to control the virtual car running on the given image. The vehicle of the team that gets to the checkpoint in the shortest time will win 9 points, the minus points will be 0.5 times for each later team.
With the excellent performance, the C213 team achieved 10 points from the jury when they completed the 5/5 checkpoint in 53.7 seconds. Next team: Rolling Thunder (9.5 points), C410 (9 points), Fteam (8.5 points), Cuda (8 points), finally Stardust and Convex Hull (7.5 points).
Within the second round of competition, teams take turns taking part in the contest: presentation – debate – answering questions from the judges in pairs. In particular, the debate section helps contestants to demonstrate their intellectual capacity, as well as create conditions for them to cultivate, discuss and learn new things in the process of answering questions from the opponent team and judges. The order of rounds is C401 – Rolling Thunder – Fteam – C213 – Cuda – Stardust – Convex Hull.
Ending with their presentation and debate, Rolling Thunder quickly overcame their rival C213 and rose to the top with three 10 points from the judges. Not less competitive, other teams such as Fteam and Cuda also won themselves the ticket to the final round with the average score of 2.83 and 7.67 respectively.
“When we use deep learning, the first important thing is data, and sometimes we don’t need to think of a completely new architecture to meet the problem, we can get those architectures on papers, and the custom work with data is already a success of deep learning”- Contestant Nguyen The Nam who came from the Cuda team criticized when receiving doubts about possible possibilities and risks when applying NVIDIA’s available algorithms.
Coming to the final round, the judges gave 3 questions with increasing difficulty and the teams have to complete the test within the prescribed time of 5 minutes for task 1, 10 minutes for task 2 and 15 minutes for task 3.
This year’s exam was rated by the contestants as more difficult than last year. However, in the blink of an eye, Rolling Thunder finished and raised the flag to signal to the judges, overwhelming the auditorium. Almost as if the victory was firmly in hand, C213 rose and overturned the situation. With an excellent, almost perfect answer, C213 convinced all judges with an absolute score and became the final winner.
After the outstanding performance in the fierce competitions, the organizers have selected two representatives for FPT University who will participate in the semifinal “Digital race” 2018-2019: C213 and Rolling Thunder.
Before that, in the 2017-2018 season, both teams from FPT University excellently won the runner-up position and the third place.
Source: FPT UniversityRelated posts: