Four years ago, organizers created the international AI City Challenge to spur the development of artificial intelligence for real-world scenarios like counting cars traveling through intersections or spotting accidents on freeways. Last week, Chinese tech giants Alibaba and Baidu swept the AI City Challenge, beating competitors from nearly 40 nations. Hundreds of Chinese cities have pilot programs, and by some estimates, China has half of the world's smart cities. One of the competitions in the AI City Challenge asked participants to identify cars in videofeeds; for the first time this year, the descriptions were in ordinary language, such as "a blue Jeep goes straight down a winding road behind a red pickup truck." He says AI researchers in the US can also compete for government grants like the National Science Foundation's Civic Innovation Challenge or the Department of Transportation's Smart City Challenge.

Four years ago, organizers created the international AI City Challenge to spur the development of artificial intelligence for real-world scenarios like counting cars traveling through intersections or spotting accidents on freeways. In the first years, teams representing American companies or universities took top spots in the competition. Last year, Chinese companies won three out of four competitions. Last week, Chinese tech giants Alibaba and Baidu swept the AI City Challenge, beating competitors from nearly 40 nations. Chinese companies or universities took first and second place in all five categories.

Vehicle re-identification aims to obtain the same vehicles from vehicle images. This is challenging but essential for analyzing and predicting traffic flow in the city. Although deep learning methods have achieved enormous progress for this task, their large data requirement is a critical shortcoming. Therefore, we propose a synthetic-to-real domain adaptation network (StRDAN) framework, which can be trained with inexpensive large-scale synthetic and real data to improve performance. The StRDAN training method combines domain adaptation and semi-supervised learning methods and their associated losses. StRDAN offers significant improvement over the baseline model, which can only be trained using real data, for VeRi and CityFlow-ReID datasets, achieving 3.1% and 12.9% improved mean average precision, respectively.