At the Vietnam Artificial Intelligence Festival (AI4VN) taking place on August 15-16, FPT brought AI autonomous car which won the Digital Race to perform at Hanoi University of Science and Technology’s courtyard.
The autonomous car contest is one of the activities organized by FPT to promote the development of the AI community. Mr. Le Ngoc Tuan – IoT Head of FPT Technology Department – Head of Arbitration in Digital Race 2018 – 2019 shared: “FPT wants to share open source DIRA for AI community in Vietnam. Also look for young people who want to attend the next Digital Race season, to build and contribute to open source in Vietnam”.
The project team The Digital Race and the 2018 – 2019 season contestants performed and shared the knowledge and source code for many participants at the event. The booths here proved a point: Students, young people can also do AI, not just doctors or professors.
Self-propelled cars are performed at a race-like event. Simulated with 1/7 size compared to standard self-propelled vehicles. With the size and high-power engine, the car can reach a maximum speed of nearly 100 km/h.
All vehicle tasks are handled through the motherboard. With AI as the “brain”, the cars are capable of handling many algorithms such as Deep Learning and Machine Learning.
For cars to run automatically, image processing technology is integrated in the front camera. Especially, the camera on the car can rotate the corners to better identify the image.
Pham Son Ha – Student of Posts and Telecommunications Institute of Technology shared: “To make this car we have applied 2 technologies, the first is Deep Learning that allows the vehicle to identify and distinguish which area is the road based on previously trained data. The second technology is the classification network technology that helps the car distinguish signs and has previously used thresholds to capture the sea on the resulting image”.
In the future, this technology can be fully applied to self-propelled robot design in a controlled space like an enterprise warehouse.
Let’s follow some pictures of autonomous cars shown in AI4VN: