HKU

On July 8th, the Hong Kong Big Data Intelligence Laboratory (HKUDS) recently released the Open Source AI Smart Body Mentor DeepTutur v. 1.5, which completed several upgrades on document decomposition engines, RAG intake pipe lines and Python 3.14+ compatibility. DeepTutor's core design is to integrate the six components of chat, testing, research, visualization, problem resolution and learning path planning into the same smart body closed loop, so students do not need to interrupt backstage engines when they switch between assignments, and learning context and multilevel memories automatically coincide..

In the hustle and bustle of the city, everyone longs for a citywalk, traveling through the streets and alleys, exploring historical sites, and immersing themselves in the local culture. However, planning a satisfying city walk is not easy. It requires the collection of travel information, the selection of points of interest (POI), route planning, and personalized customization. Now, a new technology, ItiNera, jointly developed by researchers from the University of Hong Kong, Shanghai Jiao Tong University, Massachusetts Institute of Technology, and Hong Kong Polytechnic University, is making this process simple and efficient. ItiNera is an open domain city walk…

The University of Hong Kong and Baidu jointly released the first smart city big model UrbanGPT, which has triggered a major breakthrough in the field of spatiotemporal prediction technology. The importance of spatiotemporal prediction technology is becoming increasingly prominent, focusing not only on the flow of traffic and people, but also covering multiple dimensions such as crime trends. However, due to insufficient urban data, traditional spatiotemporal prediction models are limited in terms of accurate prediction. The emergence of UrbanGPT fills this gap, and its strong generalization ability makes it show great potential in the field of zero-shot learning. Challenges faced by UrbanGPT include scarce labels and high training costs…

Recently, a research team from HKU and NYU collaborated to develop a platform called V-IRL, which incorporates rich information such as real-world maps and street views into virtual environments to provide AI Agents with a more realistic and complete life experience. This research centers on building a scalable platform to embed AI intelligences into real urban environments using APIs such as geographic information and street view images. The researchers used V-IRL to instantiate a series of intelligences that can perform a variety of tasks, such as finding restaurants, housing, and urban planning. On this platform, the intelligences can also be inter...