Lukai Li

Lukai Li

Auto-generated profile, based on publicly available information

Lukai Li

Information Engineering University, Zhengzhou, Henan, China

Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

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

Representative Publications

View All Publications
7 November 2022
Verified
Conference Paper
Open access
XYXiaokang Yin
Xiaokang Yin
RCRuijie Cai
Ruijie Cai
YZYizheng Zhang
Yizheng Zhang
LLLukai Li
Lukai Li
QYQichao Yang
Qichao Yang
SLShengli Liu
Shengli Liu
Command injection vulnerability is a severe threat to the embedded device. Most methods detect command injection vulnerability with taint analysis and symbolic execution and achieve promising results. However, they waste too much time analyzing secure sink call-sites, resulting in less efficient vulnerability detection. To tackle the above problem, we propose a novel sink call-site classification method named CINDY to accelerate the command injection vulnerability discovery in embedded firmware with static backtracking analysis. CINDY first performs sink call-sites detection in the binary executables and constructs the data flow for function call parameters. Then, CINDY analyzes whether the parameters passed to sink functions are derived from constant string or not and labels them “secure" or “risky". According to the labels, CINDY classifies the sink call-sites into risky and secure sink call-sites. Finally, CINDY performs taint analysis with symbolic execution to check whether a risky sink call-site is vulnerable. To demonstrate the efficacy of CINDY, we compare CINDY with the state-of-the-art method SaTC, using the dataset published by SaTC. Compared with SaTC, CINDY can filter out more of the secure sink call-sites, with a 35% decrease, and the efficiency is improved by 17% than SaTC.

9 August 2022 in Security and Communication Networks
Verified
Journal Article
Open access
MLMing Liu
Ming Liu
RCRuijie Cai
Ruijie Cai
LLLukai Li
Lukai Li
JWJiafeng Wang
Jiafeng Wang
QYQichao Yang
Qichao Yang
Although a large number of works have been done to explore efficient face detection in various scenes, practical face detection in unconstrained condition of varying lighting, pose, scale, and occlusion remains a challenging task. The primary limitation of existing solutions is that they are vulnerable to influence from the wild environment. Practical features extraction plays a crucial part in the face detection of low-quality images. Based on the EfficientNet, this paper builds a novel pyramid attention network to integrate multilevel features with rich context messages. Firstly, a context model is exploited to increase the receptive fields at the beginning of the network. Secondly, stacked pyramid feature attention modules and feature fusion simultaneously selectively integrate the contextual information and enable spatial details, thus enhancing the capacity to detect faces on hard images. In addition, hard samples augmentation of the training sets is conducted, which is beneficial for improving the accuracy. A thorough study on ablation verifies the effect of the proposed strategies. Moreover, extensive experiments on Wider Face and FDDB datasets, remarkably pushing the accuracy both of 96.3% (2%↑), demonstrate the performance of the proposed deep face detector which is superior and outperforms most of the preexisting methods. The method presented in this paper can perform the task of face detection in surveillance images well.

22 December 2024
Verified
Conference Paper
Open access
FXFanxiao Xing
Fanxiao Xing
LLLukai Li
Lukai Li
XLXiao Li
Xiao Li
QYQichao Yang
Qichao Yang
The widespread adoption of diverse embedded devices has made our lives more convenient, yet these devices are frequently susceptible to security flaws due to the web interfaces commonly found in their firmware. These interfaces can permit unverified input to reach critical operations, potentially leading to vulnerabilities. Static taint analysis is a proven technique for detecting and addressing such vulnerabilities. Regrettably, current static taint analysis techniques struggle to accurately and completely pinpoint the origins of contamination. To tackle this challenge, we introduce SFtaint, an optimization technique for identifying taint sources in embedded devices. Based on the relationship between the front and back ends of firmware, SFtaint leverages semantic similarity to discover more taint keywords, identifies taint import functions according to the characteristics of function data flow, and checks the validity of taint sources on the path from taint keywords to sinks. We implemented a prototype of SFtaint and evaluated its effectiveness on ten IoT devices from four vendors. Compared with the current SOTA method SaTC[1], SFtaint shows better performance in taint source identification, with an increase of 154% in taint source effectiveness.