Story Highlights
- Indian AI researchers have built the world’s first open-source, fully-controllable multi-agent driving simulator.
- The simulator has the ability to model a multi-vehicle environment closer to their real-world behaviour in different traffic scenarios.
- Support for multi-agent algorithms will enable the creation of more realistic and reactive autonomous driving behaviours, the researchers say.
IIT researchers have built the world’s first open source multi-agent driving simulator. The simulator can model a multi-vehicle environment closer to their real-world behavior in different traffic scenario.
AI researchers working on autonomous vehicles technology usually rely on simulators to mimic real-life conditions. The problem is that most of these simulators are either proprietary or inaccessible ones that Uber and Google are using. The existing open source solutionsCarla and DeepDrive control one vehicle at a time.
Abhishek Naik and Anirban Santara, the IITian duo has developed a new version of the simulator. Naik is a 22-year old graduate of IIT Madras, currently doing an internship at Parallel Computing Lab at Intel India. His fellow teammate, Santara is a 25-year old Google India PhD fellow IIT-Kharagpur, currently interning at PCL-India lab.
The team has already released the code of their MADRaS (Multi-Agent DRiving Simulator) on Github last month. The product is a multi-agent version of TORCS. It has added the ability to use control algorithms on multiple tracks running on the same track. The team took help Balaraman Ravindran, head of Robert Bosch Centre for Data Science and Artificial Intelligence at IIT-Madras.
👉 To view more please visit my blog https://steemit.com/@devrajsinghrawat
📌 Thank you in advance for your comments...
📌 Mannabase is an online platform for the world’s first Universal Basic Income cryptocurrency. You are also welcome!
https://www.mannabase.com/?ref=e7e72550c8
👉 If you believe that content delivered are with great quality then do consider to Vote as witness
⚡ All your RESTEEM's always are much appreciated here, and always noticed ⚡
🚀 Stay Tuned & Keep Steeming,
Cheers