ai-powered autonomous driving vehicle

A research team consisting of Oskar Natan, a Ph.D. student, and his supervisor, Professor Jun Miura, who are affiliated with the Active Intelligent System Laboratory (AISL), Department of Computer Science Engineering, Toyohashi University of Technology, has developed an AI model that can handle perception and control simultaneously for an autonomous driving vehicle. The AI model perceives the environment by completing several vision tasks while driving the vehicle following a sequence of route points.

Moreover, the AI model can drive the vehicle safely in diverse environmental conditions under various scenarios. Evaluated under point-to-point navigation tasks, the AI model achieves the best drivability of certain recent models in a standard simulation environment.

Autonomous driving is a complex system consisting of several subsystems that handle multiple perception and control tasks. However, deploying multiple task-specific modules is costly and inefficient, as numerous configurations are still needed to form an integrated modular system. Furthermore, the integration process can lead to information loss as many parameters are adjusted manually.

With rapid deep learning research, this issue can be tackled by training a single AI model with end-to-end and multi-task manners. Thus, the model can provide navigational controls solely based on the observations provided by a set of sensors. As manual configuration is no longer needed, the model can manage the information all by itself.

The challenge that remains for an end-to-end model is how to extract useful information so that the controller can estimate the navigational controls properly. This can be solved by providing a lot of data to the perception module to better perceive the surrounding environment. In addition, a sensor fusion technique can be used to enhance performance as it fuses different sensors to capture various data aspects.

However, a huge computation load is inevitable as a bigger model is needed to process more data. Moreover, a data preprocessing technique is necessary as varying sensors often come with different data modalities. Furthermore, imbalance learning during the training process could be another issue since the model performs both perception and control tasks simultaneously.

In order to answer those challenges, the team propose an AI model trained with end-to-end and multi-task manners. The model is made of two main modules, namely perception and controller modules. The perception phase begins by processing RGB images and depth maps provided by a single RGBD camera. Then, the information extracted from the perception module along with vehicle speed measurement and route point coordinates are decoded by the controller module to estimate the navigational controls.

So as to ensure that all tasks can be performed equally, the team employs an algorithm called modified gradient normalization (MGN) to balance the learning signal during the training process. The team considers imitation learning as it allows the model to learn from a large-scale dataset to match a near-human standard. Furthermore, the team designed the model to use a smaller number of parameters than others to reduce the computational load and accelerate the inference on a device with limited resources.

Based on the experimental result in a standard autonomous driving simulator, CARLA, it is revealed that fusing RGB images and depth maps to form a birds-eye-view (BEV) semantic map can boost the overall performance. As the perception module has better overall understanding of the scene, the controller module can leverage useful information to estimate the navigational controls properly. Furthermore, the team states that the proposed model is preferable for deployment as it achieves better drivability with fewer parameters than other models.

Future outlook
The team is currently working on modifications and improvements to the model so as to tackle several issues when driving in poor illumination conditions, such as at night, in heavy rain, etc. As a hypothesis, the team believes that adding a sensor that is unaffected by changes in brightness or illumination, such as LiDAR, will improve the model’s scene understanding capabilities and result in better drivability. Another future task is to apply the proposed model to autonomous driving in the real world.

Research Report:End-to-end Autonomous Driving with Semantic Depth Cloud Mapping and Multi-agent

Keyword: AI-powered autonomous driving vehicle


Best cars for a 'Love Bug' remake

Tennessee’s Yard Art The Thorndyke Special The Hot Rod The Hippie Van Jim’s Lamborghini Growing up, it’s safe to say my absolute favorite movie was Disney’s “The Love Bug.” As a kid living in a world before Pixar’s “Cars,” it was pretty much the best car movie. I loved ...

View more: Best cars for a 'Love Bug' remake

Audi timing belt tensioner prompts Utah bomb squad visit

An aftermarket timing belt tensioner found in a Salt Lake City apartment prompted a visit from the bomb squad this week after it was mistaken for a potential explosive device. Police partially evacuated the apartment complex in which it was found “out of an abundance of caution” and called ...

View more: Audi timing belt tensioner prompts Utah bomb squad visit

2023 GMC Sierra HD 2500 and 3500 will get more expensive

GMC Sierra HD pickups are about to get more expensive again after recently becoming more expensive, according to GM Authority. In July, GMC and Buick raised the prices on certain vehicles in their lineups anywhere from $975 to $1,675 by making the previously optional OnStar a standard feature. Next ...

View more: 2023 GMC Sierra HD 2500 and 3500 will get more expensive

Our week with EVs: Recapping the diverse collection of electric cars we tested

Mercedes-Benz EQB First Drive Review: Next electric Benz is a little different 2022 Chevy Bolt EV Road Test Review: Time to play EV, gas or airplane 2023 Genesis GV60 Road Test Review: The third of Hyundai Group’s E-GMP EVs is the fastest, but softest It may be hard to ...

View more: Our week with EVs: Recapping the diverse collection of electric cars we tested

Europe car sales lowest since 1996 after 12-month decline

BERLIN – Europe registered the lowest number of new passenger cars in the month of June this year since 1996 at just over 1.06 million vehicles, with some carmakers seeing sales drop by nearly 50%, data from Europe’s automobile association showed on Friday. Volkswagen Group was the hardest-hit major ...

View more: Europe car sales lowest since 1996 after 12-month decline

More Than Just A Carmaker: Toyota Motor PH Launches Toyota Mobility Solutions

Automotive industry leader, Toyota Motor Philippines Corporation (TMP) inaugurated today Toyota Mobility Solutions Philippines, Inc. (TMSPH), a wholly owned subsidiary that will be a provider of mobility-related services. TMSPH ventures into the development and offering of a range of ‘new mobility solutions’ that will enhance Toyota as a brand ...

View more: More Than Just A Carmaker: Toyota Motor PH Launches Toyota Mobility Solutions

'F1 22' feels fast and familiar | Gaming Roundup

‘F1 22’ impressions ‘Construction Simulator’ is launching Sept. 20 Autoblog may receive a share from purchases made via links on this page. Pricing and availability are subject to change. This week in racing game news: ‘F1 22’ impressions At its core, “F1 22,” the latest installment in the F1 ...

View more: 'F1 22' feels fast and familiar | Gaming Roundup

Watch a single-engine plane crash-land on 91 Freeway in California

A single-engine Piper Cherokee airplane was forced to crash land on the 91 Freeway in California after losing power about 45 minutes outside of Los Angeles. Yes, that means this was a very busy highway, practically bursting at the seams with traffic. No, amazingly nobody died, either in the ...

View more: Watch a single-engine plane crash-land on 91 Freeway in California

Ford files 'Mustang Dark Horse' trademark application

Make adventures more comfortable with high-quality Jeep grab handles

New Lucid Air variant to debut & Stealth Look to be on display during Monetary Car Week

Chevy offers incentives to prevent Corvette Z06 flipping

Average U.S. gas price falls below $4/gallon

German court: SUV driver must pay more than a car for running red light

Australian vehicle crash tests to include underwater performance

Toyota C-HR rumored to get revamp next year and an EV version

Geely's electric Zeekr 009 is an EV Alphard we never had; 700 km, 542 PS, top speed 190 km/h

A Newey, Brawn and Anderson row that shows F1 doesn’t change

Sorry, EV buyers, you won't get to pick your own pedestrian safety sounds

There won't be enough copper to meet climate goals, study indicates


; Tin công nghệ, điện thoại, máy tính, ô tô, phân khối lớn, xu hướng công nghệ cập nhật mới nhất