Lionel M. Ni

Lionel M. Ni

Auto-generated profile, based on publicly available information

Lionel M. Ni

The Hong Kong University of Science and Technology (Guangzhou), Guangzhou, China

Research Focus & Metrics

Academic Career Length

35Years (1992-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 April 2024 in IEEE Transactions on Pattern Analysis and Machine Intelligence
Verified
Conference Paper
Open access
FLFeng Li
Feng Li
HZHao Zhang
Hao Zhang
SLShilong Liu
Shilong Liu
JGJian Guo
Jian Guo
LNLionel M. Ni
Lionel M. Ni
LZLei Zhang
Lei Zhang
We present in this paper a novel denoising training method to speedup DETR (DEtection TRansformer) training and offer a deepened understanding of the slow convergence issue of DETR-like methods. We show that the slow convergence results from the instability of bipartite graph matching which causes inconsistent optimization goals in early training stages. To address this issue, except for the Hungarian loss, our method additionally feeds ground-truth bounding boxes with noises into Transformer decoder and trains the model to reconstruct the original boxes, which effectively reduces the bipartite graph matching difficulty and leads to a faster convergence. Our method is universal and can be easily plugged into any DETR-like methods by adding dozens of lines of code to achieve a remarkable improvement. As a result, our DN-DETR results in a remarkable improvement (+1.9AP) under the same setting and achieves the best result (AP 43.4 and 48.6 with 12 and 50 epochs of training respectively) among DETR-like methods with ResNet-50 backbone. Compared with the baseline under the same setting, DN-DETR achieves comparable performance with 50% training epochs. Code is available at https://github.com/FengLi-ust/DN-DETR.

1 January 1992
Verified
Conference Paper
Open access
CGChristopher J. Glass
Christopher J. Glass
LNLionel M. Ni
Lionel M. Ni

25 August 2020 in IEEE Internet of Things Journal
Verified
Journal Article
Closed access
YDYe Ding
Ye Ding
WZWenyi Zhang
Wenyi Zhang
XZXibo Zhou
Xibo Zhou
QLQing Liao
Qing Liao
QLQiong Luo
Qiong Luo
LNLionel M. Ni
Lionel M. Ni
A passenger is overcharged by the taxi driver is one common type of fraudulent trip, and it brings negative impacts to modern cities. Most existing fraudulent trip detection works rely on the assumption that the trip is correctly recorded by the taximeter. However, there are many taxi drivers in China carrying passengers without activating the taximeter, especially when the taxi driver is trying to overcharge the passengers. Hence, existing detection methods cannot be directly applied to such real-world scenario. In this paper, we propose a system, called “FraudTrip”, which detects “unmetered” taxi trips based on a novel fraud detection algorithm and a heuristic maximum fraudulent trajectory construction algorithm. Based on the experiments on both synthetic and real-world trajectory datasets, FraudTrip can effectively and efficiently detect fraudulent trips without the help of taximeters.

30 May 2018 in IEEE Transactions on Image Processing
Verified
Journal Article
Closed access
YWYaqing Wang
Yaqing Wang
QYQuanming Yao
Quanming Yao
JKJames Kwok
James Kwok
LNLionel M. Ni
Lionel M. Ni
Convolutional sparse coding (CSC) improves sparse coding by learning a shift-invariant dictionary from the data. However, most existing CSC algorithms operate in the batch mode and are expensive, in terms of both space and time, on large data sets. In this paper, we alleviate these problems by using online learning. The key is a reformulation of the CSC objective so that convolution can be handled easily in the frequency domain and much smaller history matrices are needed. We use the alternating direction method of multipliers (ADMM) to solve the resultant optimization problem, and the ADMM subproblems have efficient closed-form solutions. Theoretical analysis shows that the learned dictionary converges to a stationary point of the optimization problem. Extensive experiments are performed on both standard CSC benchmark data sets and much larger data sets such as the ImageNet. Results show that the proposed algorithm outperforms state-of-the-art batch and online CSC methods by being more scalable, and has faster convergence and better reconstruction performance.

1 January 2025
Verified
Conference Paper
Open access
PDF file
SWSaizhuo Wang
Saizhuo Wang
HYHang Yuan
Hang Yuan
LZLeon Zhou
Leon Zhou
LNLionel Ni
Lionel Ni
HSHeung-Yeung Shum
Heung-Yeung Shum
JGJian Guo
Jian Guo
Saizhuo Wang, Hang Yuan, Leon Zhou, Lionel Ni, Heung-Yeung Shum, Jian Guo. Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing: System Demonstrations. 2025.

31 January 2023 in Journal of Computer Science and Technology
Verified
Journal Article
Open access
PDF file
JHJia-Ling Huang
Jia-Ling Huang
YWYun-Shu Wang
Yun-Shu Wang
YZYong-Pan Zou
Yong-Pan Zou
KWKaishun Wu
Kaishun Wu
LNLionel Ming-Shuan Ni
Lionel Ming-Shuan Ni
With the increasing pervasiveness of mobile devices such as smartphones, smart TVs, and wearables, smart sensing, transforming the physical world into digital information based on various sensing medias, has drawn researchers’ great attention. Among different sensing medias, WiFi and acoustic signals stand out due to their ubiquity and zero hardware cost. Based on different basic principles, researchers have proposed different technologies for sensing applications with WiFi and acoustic signals covering human activity recognition, motion tracking, indoor localization, health monitoring, and the like. To enable readers to get a comprehensive understanding of ubiquitous wireless sensing, we conduct a survey of existing work to introduce their underlying principles, proposed technologies, and practical applications. Besides we also discuss some open issues of this research area. Our survey reals that as a promising research direction, WiFi and acoustic sensing technologies can bring about fancy applications, but still have limitations in hardware restriction, robustness, and applicability.