Autonomous valet parking is a specific application for autonomous vehicles. In this task, vehicles need to navigate in narrow, crowded and GPS-denied lots. Accurate localization ability of great importance. Traditional visual-based methods suffer from tracking lost due texture-less regions, repeated structures, appearance changes. paper, we exploit robust semantic features build the map localize Semantic contain guide signs, lines, speed bumps, etc, which typically appear Compared with...

Accurate localization is of crucial importance for autonomous driving tasks. Nowadays, we have seen a lot sensor-rich vehicles (e.g. Robo-taxi) on the street autonomously, which rely high-accurate sensors Lidar and RTK GPS) high-resolution map. However, low-cost production cars cannot afford such high expenses maps. How to reduce costs? do benefit cars? In this paper, proposed light-weight solution, relies cameras compact visual semantic The map easily produced updated by in crowd-sourced...

Lithium-ion batteries with high energy density and good safety are in urgent demand nowadays. Silicon has the advantages of theoretical capacity, low cost, vast reserves. It is regarded as an ideal anode material for high-energy lithium-ion batteries. Unfortunately, its huge volume change during cycling results unstable cycle behavior limits application. To develop a robust electrode by employing effective binders still big challenge. Furthermore, both conventional organic liquid...

In order to study the mechanics and acoustic emission (AE) characteristics of fractured rock under water-rock interaction, dried saturated sandstone samples with prefabricated double parallel cracks were prepared. Then, uniaxial compression experiments performed obtain their AE signals crack propagation images. The results show that water reduces strength fracture toughness enhances plasticity. After saturation, start earlier; grow slowly; failure mode is transformed from shear along...

An accurate road topological structure is of great importance for autonomous driving in complex urban environments. Currently, most vehicles highly rely on the High-Definition map (HD map) to cruise across city. Without prior map, it's hard find right-turning and left-turning ways large intersections. However, due complexity intersections, producing such a by human resources time-consuming error-prone. In this letter, we proposed framework automatically produce complicated This adopts...

Accurate localization is of crucial importance for autonomous driving tasks. Nowadays, we have seen a lot sensor-rich vehicles (e.g. Robo-taxi) on the street autonomously, which rely high-accurate sensors Lidar and RTK GPS) high-resolution map. However, low-cost production cars cannot afford such high expenses maps. How to reduce costs? do benefit cars? In this paper, proposed light-weight solution, relies cameras compact visual semantic The map easily produced updated by in crowd-sourced...

Autonomous valet parking is a specific application for autonomous vehicles. In this task, vehicles need to navigate in narrow, crowded and GPS-denied lots. Accurate localization ability of great importance. Traditional visual-based methods suffer from tracking lost due texture-less regions, repeated structures, appearance changes. paper, we exploit robust semantic features build the map localize Semantic contain guide signs, lines, speed bumps, etc, which typically appear Compared with...

Sensing environmental obstacles and establishing an occupancy map of surroundings are critical to achieving automated parking for autonomous vehicles. This paper presents a method obtain surrounding information from inverse perspective mapping (IPM) images. uses the easily-accessed pseudo-labels LiDAR supervise visual network, which can detect occupied boundaries obstacles. Fusing this with ego-motion information, we develop multi-frame fusion approach build local OGM realize online...

Entity alignment is an important task in knowledge fusion, which aims to link entities that have the same real-world identity two graphs. However, process of constructing a graph, some noise may inevitably be introduced, must affect results entity tasks. The triple confidence calculation can quantify correctness triples reduce impact on alignment. Therefore, we designed method calculate and applied it representation learning phase calculates based pairing rates three angles between...

Mining maximal frequent itemsets is an important research topic in data mining. The existing FPMAX algorithm for mining based on FP-tree necessitate to construct conditional subset trees recursively. While the amount of large, it will cause huge consumption both terms runtime and memory. Therefore, this paper proposes improved FPMAX-direct FP-tree. This novel simplify construction by adding directly branch whose support no less than a user-specified minimum threshold candidate itemsets....

Compared with other neural networks, Radial Basis Function (RBF) network has the advantages of simple structure and fast convergence. As long as there are enough hidden layer nodes in layer, it can approximate any non-linear function. In this paper, finite element model a through tied arch bridge is modified based on Neural Network. The approximation function RBF utilized to fit implicit relationship between response its design parameters. Then modified. results show that efficient modify bridge.