CAS's dynamic topology perception empowers embodied navigation! DGNav: breaking through the granular rigidity problem of visual-language navigation

Posted: 2026/3/16 | Category: Artificial Intelligence
by Jiankun Peng, Jianyuan Guo, Ying Xu, Yue Liu, Jiashuang Yan, Xuanwei Ye, Houhua Li, Xiaoming Wang | with Institute of Aerospace Information Innovation, Chinese Academy of Sciences, School of Electronics, Electrical Engineering and Communication, University of Chinese Academy of Sciences, Department of Computer Science, City University of Hong Kong

Visual-Language Navigation (VLN) is a core task of embodied intelligence, requiring an intelligent agent to interpret natural language instructions and navigate to a target location in an unknown environment.

+ Read More
40 %

Photonic Simulation of Majorana-Based Jones Polynomials

Posted: 2025/1/6 | Category: Artificial Intelligence
03/2021 - 12/2024 | by 郭光灿 教授, Prof Jiannis Pachos | with China University of Science and Technology, University of Leeds

By braiding non-Abelian anyons it is possible to realize fault-tolerant quantum algorithms through the computation of Jones polynomials. So far, this has been an experimentally formidable task. In this Letter, a photonic quantum system employing two-photon correlations and nondissipative imaginary-time evolution is utilized to simulate two inequivalent braiding operations of Majorana zero modes. The resulting amplitudes are shown to be mathematically equivalent to Jones polynomials. The high fidelity of our optical platform allows us to distinguish between a wide range of links, such as Hopf links, Solomon links, Trefoil knots, Figure Eight knots and Borromean rings, through determining their corresponding Jones polynomials. Our photonic quantum simulator represents a significant step towards executing fault-tolerant quantum algorithms based on topological quantum encoding and manipulation.

+ Read More
100 %

Highlight Projects