Search

Your search keyword '"Li, Xianfeng"' showing total 2,315 results
Start Over Author "Li, Xianfeng"
2,315 results on '"Li, Xianfeng"'

1. GAD-Generative Learning for HD Map-Free Autonomous Driving

Author

Sun, Weijian, Jia, Yanbo, Zeng, Qi, Liu, Zihao, Liao, Jiang, Li, Yue, and Li, Xianfeng

Subjects
Computer Science - Robotics, Computer Science - Computer Vision and Pattern Recognition
Abstract

Deep-learning-based techniques have been widely adopted for autonomous driving software stacks for mass production in recent years, focusing primarily on perception modules, with some work extending this method to prediction modules. However, the downstream planning and control modules are still designed with hefty handcrafted rules, dominated by optimization-based methods such as quadratic programming or model predictive control. This results in a performance bottleneck for autonomous driving systems in that corner cases simply cannot be solved by enumerating hand-crafted rules. We present a deep-learning-based approach that brings prediction, decision, and planning modules together with the attempt to overcome the rule-based methods' deficiency in real-world applications of autonomous driving, especially for urban scenes. The DNN model we proposed is solely trained with 10 hours of human driver data, and it supports all mass-production ADAS features available on the market to date. This method is deployed onto a Jiyue test car with no modification to its factory-ready sensor set and compute platform. the feasibility, usability, and commercial potential are demonstrated in this article.

Published
2024

6. Genomic characterization and risk stratification of esophageal squamous dysplasia

Author

Min Qingjie, Zhang Min, Lin Dongmei, Zhang Weimin, Li Xianfeng, Zhao Lianmei, Teng Huajing, He Tao, Sun Wei, Fan Jiawen, Yu Xiying, Chen Jie, Li Jinting, Gao Xiaohan, Dong Bin, Liu Rui, Liu Xuefeng, Song Yongmei, Cui Yongping, Lu Shih-Hsin, Li Ruiqiang, Guo Mingzhou, Wang Yan, and Zhan Qimin

Subjects
esophageal squamous dysplasia, genomic alteration, acsm5, early diagnosis, mir-4292, Medicine
Abstract

The majority of esophageal squamous dysplasia (ESD) patients progress slowly, while a subset of patients can undergo recurrence rapidly or progress to invasive cancer even after proper treatment. However, the molecular mechanisms underlying these clinical observations are still largely unknown.

Published
2024
Full Text
View/download PDF

7. A joint real-time SOC estimation method for energy storage batteries

Author

LI Xianfeng, HU Chengang, BU Liming, CHEN Pan, MIAO Wenjie, and HUANG Wenzhe

Subjects
soc, vffrls, ukf, Electrical engineering. Electronics. Nuclear engineering, TK1-9971
Abstract

Accurately estimating the state of charge （SOC） of energy storage batteries is of paramount importance for achieving balanced charging and discharging， and mitigating capacity degradation caused by overcharging and overdischarging. In view of the complex chemical states and nonlinear time-varying characteristics of SOC in energy storage batteries， this paper proposes a joint SOC estimation method for lithium-ion batteries based on the variable forgetting factor recursive least squares （VFFRLS） and unscented Kalman filter （UKF） algorithms. The VFFRLS algorithm is employed for online identification of battery model parameters such as resistance and capacitance， and based on the identification results， the UKF algorithm is utilized for real-time SOC estimation. Experimental results demonstrate that the proposed joint method exhibits high accuracy and stability.

Published
2024
Full Text
View/download PDF

9. Research on Homography Estimation Method Based on Deep Learning

Author

Shan, Chun, Li, Xianfeng, Chen, Xingchi, Zhou, Houlin, Zou, Tongyi, Li, Jiayi, Liu, Shaoming, Ceccarelli, Marco, Series Editor, Corves, Burkhard, Advisory Editor, Glazunov, Victor, Advisory Editor, Hernández, Alfonso, Advisory Editor, Huang, Tian, Advisory Editor, Jauregui Correa, Juan Carlos, Advisory Editor, Takeda, Yukio, Advisory Editor, Agrawal, Sunil K., Advisory Editor, and Zhou, Kun, editor

Published
2024
Full Text
View/download PDF

10. Attention Diversification for Domain Generalization

Author

Meng, Rang, Li, Xianfeng, Chen, Weijie, Yang, Shicai, Song, Jie, Wang, Xinchao, Zhang, Lei, Song, Mingli, Xie, Di, and Pu, Shiliang

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

Convolutional neural networks (CNNs) have demonstrated gratifying results at learning discriminative features. However, when applied to unseen domains, state-of-the-art models are usually prone to errors due to domain shift. After investigating this issue from the perspective of shortcut learning, we find the devils lie in the fact that models trained on different domains merely bias to different domain-specific features yet overlook diverse task-related features. Under this guidance, a novel Attention Diversification framework is proposed, in which Intra-Model and Inter-Model Attention Diversification Regularization are collaborated to reassign appropriate attention to diverse task-related features. Briefly, Intra-Model Attention Diversification Regularization is equipped on the high-level feature maps to achieve in-channel discrimination and cross-channel diversification via forcing different channels to pay their most salient attention to different spatial locations. Besides, Inter-Model Attention Diversification Regularization is proposed to further provide task-related attention diversification and domain-related attention suppression, which is a paradigm of "simulate, divide and assemble": simulate domain shift via exploiting multiple domain-specific models, divide attention maps into task-related and domain-related groups, and assemble them within each group respectively to execute regularization. Extensive experiments and analyses are conducted on various benchmarks to demonstrate that our method achieves state-of-the-art performance over other competing methods. Code is available at https://github.com/hikvision-research/DomainGeneralization., Comment: ECCV 2022. Code available at https://github.com/hikvision-research/DomainGeneralization

Published
2022

11. Roadmap for rechargeable batteries: present and beyond

Author

Xin, Sen, Zhang, Xu, Wang, Lin, Yu, Haijun, Chang, Xin, Zhao, Yu-Ming, Meng, Qinghai, Xu, Pan, Zhao, Chen-Zi, Chen, Jiahang, Lu, Huichao, Kong, Xirui, Wang, Jiulin, Chen, Kai, Huang, Gang, Zhang, Xinbo, Su, Yu, Xiao, Yao, Chou, Shu-Lei, Zhang, Shilin, Guo, Zaiping, Du, Aobing, Cui, Guanglei, Yang, Gaojing, Zhao, Qing, Dong, Liubing, Zhou, Dong, Kang, Feiyu, Hong, Hu, Zhi, Chunyi, Yuan, Zhizhang, Li, Xianfeng, Mo, Yifei, Zhu, Yizhou, Yu, Dongfang, Lei, Xincheng, Zhao, Jianxiong, Wang, Jiayi, Su, Dong, Guo, Yu-Guo, Zhang, Qiang, Chen, Jun, and Wan, Li-Jun

Published
2024
Full Text
View/download PDF

15. An operational strategy for residential microgrids considering emergency power source functionality

Author

LI Xianfeng, HU Chengang, BU Limin, MIAO Wenjie, HUANG Wenzhe, and MA Jianjun

Subjects
residential microgrid, emergency power source functionality, demand-side response, elastic coefficient, risk coefficient, energy management, Electrical engineering. Electronics. Nuclear engineering, TK1-9971
Abstract

As the societal demand for electricity rises and extreme weather events become more frequent， power interruptions are occurring， leading to a decline in the quality of electricity for users such as residential communities. In response to this issue， a study was conducted on the operational strategies of residential microgrids incorporating emergency power sources. Initially， a model for a residential microgrid incorporating emergency power sources was established， including distributed PV， energy storage， electric vehicles， and other control resources. Subsequently， in regular conditions， a risk assessment and energy pre-management strategy was proposed with the objective of minimizing electricity purchase costs and constraining energy storage. In the event of external power supply interruptions， while ensuring the normal supply of essential loads， an emergency energy management strategy was introduced with the goals of maximizing resource utilization and minimizing operational costs， using energy storage capacity as a constraint. Finally， the effectiveness of the proposed operational strategies was validated through simulation experiments simulating load interruptions in residential communities. This operational strategy effectively addresses power interruptions in residential communities， balancing the security and reliability of residential power supply while considering the utilization rate of emergency power resources and economic operational benefits.

Published
2024
Full Text
View/download PDF

38. A Free Lunch for Unsupervised Domain Adaptive Object Detection without Source Data

Author

Li, Xianfeng, Chen, Weijie, Xie, Di, Yang, Shicai, Yuan, Peng, Pu, Shiliang, and Zhuang, Yueting

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

Unsupervised domain adaptation (UDA) assumes that source and target domain data are freely available and usually trained together to reduce the domain gap. However, considering the data privacy and the inefficiency of data transmission, it is impractical in real scenarios. Hence, it draws our eyes to optimize the network in the target domain without accessing labeled source data. To explore this direction in object detection, for the first time, we propose a source data-free domain adaptive object detection (SFOD) framework via modeling it into a problem of learning with noisy labels. Generally, a straightforward method is to leverage the pre-trained network from the source domain to generate the pseudo labels for target domain optimization. However, it is difficult to evaluate the quality of pseudo labels since no labels are available in target domain. In this paper, self-entropy descent (SED) is a metric proposed to search an appropriate confidence threshold for reliable pseudo label generation without using any handcrafted labels. Nonetheless, completely clean labels are still unattainable. After a thorough experimental analysis, false negatives are found to dominate in the generated noisy labels. Undoubtedly, false negatives mining is helpful for performance improvement, and we ease it to false negatives simulation through data augmentation like Mosaic. Extensive experiments conducted in four representative adaptation tasks have demonstrated that the proposed framework can easily achieve state-of-the-art performance. From another view, it also reminds the UDA community that the labeled source data are not fully exploited in the existing methods., Comment: accepted by AAAI2021

Published
2020

Catalog

Books, media, physical & digital resources
See catalog results