Chinese AI Startup TuSimple Breaks Ten Records in Autonomous-driving Technology

SAN DIEGO, Oct. 12, 2016 /PRNewswire/ -- TuSimple, a Chinese computer vision and artificial intelligence startup, announced that it ranked No. 1 in KITTI and Cityscapes, the most influential public leaderboard in autonomous driving.

For KITTI, TuSimple swept three records in object detection, two in object tracking and four in road segmentation. In total, TuSimple achieved world-leading results in 10 records.

KITTI/CityScapes dataset has been a popular arena for many years. Its players include many world-class research institutes, such as Baidu, Samsung, NVidia, and NEC, and top universities, such as Stanford, and University of California.

An authoritative public benchmark dataset is important to evaluate the technical competence of a team. The KITTI Vision Benchmark Suite, established by Karlsruhe Institute of Technology and Toyota Technological Institute at Chicago, is the world's first and largest benchmark for vision based autonomous driving. KITTI includes real images collected from a variety of road scenes, from urban streets to country roads to highways. Each image contains a sophisticated scenario involving, for instance, a crowded vehicle and pedestrians, with various levels of occlusion.

KITTI object detection includes vehicle, pedestrian and bicycle detection. KITTI target tracking includes vehicle and pedestrian tracking. KITTI road segmentation includes four individual scenarios, including urban unmarked, urban marked, urban multiple marked and the average of former named urban road.

TuSimple swept KITTI's nine individual tests, ranking first in the world for all of them, while other well-known institutes had previously had only one or two individual top ranks.

Cityscapes Dataset is published by Mercedes-Benz and provides a segmentation data set in anonymous driving. It is used to evaluate algorithms' performance of semantic understanding in an urban setting. Cityscapes have 50 cities with different scenes, backgrounds and seasons. It has 5,000 fine annotation images, 20,000 roughly annotation images and 30 class objects.

Cityscapes benchmark has two subsets: fine and coarse. The former provides 5,000 very detailed, pixel-level labeling and the latter provides an extra 20,000 coarse level labeling. TuSimple's algorithm triumphed under each sets of criteria.

In addition to TuSimple's success in the self-driving benchmark for KITTI and Cityscapes, TuSimple also achieved first place in facial landmark localization benchmark, 300W and AFLW by a landslide. This technique is mainly used for driver monitoring systems and positioning driver facial landmarks to detect fatigue or distracted driving.

The same technologies have been used in TuSimple's product and demo.

About TuSimple

TuSimple is a computer vision and artificial intelligence algorithms technology provider that primarily helps domestic enterprises to customize the image recognition, autonomous driving, advanced driver assistance system and driver monitoring system technologies using computer vision and deep learning algorithms.

As a technology-driven artificial intelligence startup, TuSimple established two R&D centers in Beijing and San Diego to hire top scientists in both U.S. and China. TuSimple team members come from California Institute of Technology, Carnegie Mellon University, the Hong Kong University of Science and Technology, Nanyang Technological University, Tsinghua University, Peking University and Shanghai Jiaotong University. Sixty percent of its employees have a Ph.D. degree.

TuSimple CTO Xiaodi Hou obtained his Ph. D. degree from Caltech in 2014. His spectral saliency theory is now the most influential research in the field of visual attention mechanisms from the past 10 years.. Principal Scientist Naiyan Wang, Ph.D., is one of the selected candidates from HKUST's 2014 Google PhD Fellow program. Not only has Wang demonstrated excellent achievements in data mining and computer vision competition, but he is also the first person in the world to apply deep learning to object tracking fields.

They both lead an algorithm team in San Diego and Beijing, respectively. TuSimple swept the two authoritative, autonomous driving rankings field, and both teams have a great contribution.

TuSimple's business model is comparable to Otto, an US autonomous truck company, which provides truck companies with an autonomous driving system for highways: it primarily collaborates with auto transportation operators customizing camera and LiDAR-based low-cost autonomous driving algorithms and solutions.

For more information visit http://www.tusimple.com/index_en.html

SOURCE TuSimple LLC