Pengfei Zhou
Research Fellow, Alibaba-NTU Joint Research Institute (JRI), Singapore
Technical Advisor, Alibaba Group, China
Email: pengfei.zhou@ntu.edu.sg
|
|
Pengfei Zhou
Research Fellow, Alibaba-NTU Joint Research Institute (JRI), Singapore
Technical Advisor, Alibaba Group, China
Email: pengfei.zhou@ntu.edu.sg
|
Biography
Pengfei Zhou is a Research Fellow in Alibaba-NTU Joint Research Institute (JRI), Singapore. He is also a Technical Advisor for Alibaba Group, adopting research technology to Ele.me (2nd largest on-demand food delivery platform in China) for courier status tracking. He received the B.E. degree in Department of Automation from Tsinghua University in 2009, and the PhD degree in School of Computer Science and Engineering from Nanyang Technological University in 2016, under the supervision of Prof. Mo Li. Before joining JRI, he founded and led fayfans Co., Ltd in Beijing developing products and services using advanced AIoT and HCI technology from 2016 to 2019.
Research Interests
Mobile and intelligent systems, Human Computer Interaction (HCI), Artificial Intelligence of Things (AIoT), 5G networking and applications.
Publication
In Submission
Experience: Nationwide Deployment of Indoor Outdoor Detection for On-demand Food Delivery Business.Conference Papers
[SenSys'21] "LIMU-BERT: Unleashing the Potential of Unlabeled Data for IMU Sensing Applications", In ACM SenSys, Coimbra, Portugal, November 15-17, 2021. (PDF), (BibTex), (Source code). Huatao Xu, Pengfei Zhou, Rui Tan, Mo Li, Guobin Shen. [Ubicomp'19] "Amateur: Augmented Reality based Vehicle Navigation System", In ACM Ubicomp, Vol.2, No.4, Article 155, London, UK, December 2019. (PDF), (BibTex), (Demo video).Journal Articles
[TOSN]
"Memento: An Emotion Driven Lifelogging System with Wearables",
In ACM Transactions on Sensor Networks, Vol.15, Issue 1, Article 8, January 2019.
Shiqi Jiang, Pengfei Zhou, Zhenjiang Li, and Mo Li.
[TMC]
"Think Like A Graph: Real-Time Traffic Estimation at City-Scale",
In IEEE Transactions on Mobile Computing, Vol. 18, Issue 10, October 2019.
Zhidan Liu, Pengfei Zhou, Zhenjiang Li, and Mo Li.
(PDF),
(BibTex).
[TMC]
"Pricing Data Tampering in Automated Fare Collection with NFC-equipped Smartphones",
In IEEE Transactions on Mobile Computing, Vol. 18, Issue 5, May 2019.
Fan Dang, Ennan Zhai, Zhenhua Li, Pengfei Zhou, Aziz Mohaisen, Kaigui Bian, Qingfu Wen, and Mo Li.
[TMC]
"An Acoustic-based Encounter Profiling System",
In IEEE Transactions on Mobile Computing, Vol 17, Issue 8, Pages 1750-1763, August 2018.
Huanle Zhang, Wan Du, Pengfei Zhou, Mo Li, and Prasant Mohapatra.
[T-ITS]
"A Participatory Urban Traffic Monitoring System: The Power of Bus Riders",
In IEEE Transactions on Intelligent Transportation Systems, Vol.18, Issue 10, October 2017.
Zhidan Liu, Shiqi Jiang, Pengfei Zhou, and Mo Li.
[TOSN]
"IODetector: A Generic Service for Indoor/Outdoor Detection",
In ACM Transactions on Sensor Networks, Vol. 11, Issue 2, Article 28, February 2015.
Mo Li, Pengfei Zhou, Yuanqing Zheng, Zhenjiang Li, and Guobin (Jacky) Shen.
[TMC]
"How Long to Wait?: Predicting Bus Arrival Time with Mobile Phone based Participatory Sensing",
In IEEE Transactions on Mobile Computing, Vol. 13, Issue 6, Pages 1228-1241, June 2014.
Pengfei Zhou, Yuanqing Zheng, and Mo Li.
Research Projects and Technology Transfer to Products
☞ [Project] Improving IoT sensing with advanced AI techniques and intelligent human-IoT interactions (2020-present)
We use advanced AI techniques like BERT and Federated Learning to solve practical challenges in large-scale sensing systems and IoT applications. We are adopting BERT from NLP domain to extract general features from massive unlabelled data in real world. The design has been validated in various open-source datasets with average accuracy over 90% for IMU-based activity recognition application.☞ [Project] Improving real time video streaming in 5G networks (2019-present)
This project aims to improve the real time video streaming performance in 5G by incorporating the video context information and highly dynamic cellular network conditions. Empirical experiments have validated our hypothesis on the problem.☞ [Product] YuFeiMen: Smart authentication system with various interfaces (2017-2019)
YuFeiMen is an intelligent identity authentication product, supporting various authentication interfaces including gesture-based commands, face recognition, dynamic QR code and NFC-based communication. The product is being used in many scenarios like office buildings, industry parks, chain hotels, etc. PoE (Power over Ethernet) technology is used in the product so that the installation and maintenance overhead is greatly reduced compared to traditional systems.
|
|
|
☞ [Project→Product] Ensuring system security for NFC-based AFC systems (2016-2019)
This project investigated the security problems and possible attacks in current NFC-based Auto Fare Collection (AFC) systems, which are widely used in the public transit systems worldwide. Based on the results, we launched the Kakachong product in 2016, which is a mobile app for toping up smart cards using NFC-equipped smartphones.
|
|
|
☞ [Project] Instant 3D orientation estimation of mobile phones for mobile systems and gaming interfaces (2013-2015)
3D phone orientation is valuable for various applications including mobile systems and orientation based HCI for mobile gaming (demo video). This project solved the problem of instant 3D orientation estimation of smartphones. We understood the characteristics of gyroscope, accelerometer and magnetometer sensor through comprehensive experiments and analysis. We proposed to continuously track the orientation primarily using gyroscope and opportunistically calibrate the orientation using accelerometer and compass when we sense the "good moment".
|
|
|
☞ [Project→Product] IODetector: Generic indoor/outdoor detection system (2011- present)
Try IODetector source code for Android devices for your purpose.
The indoor/outdoor status is the essential and primitive information for various applications. We proposed IODetector, which tracks people's indoor/outdoor status using the smartphone on-board sensors (e.g., light sensor, cellular module and magnetism sensor, etc.) and achieves high accuracy with little power overhead. The pilot prototype was developed in 2012. In 2019, we did a technology transfer of IODetector for Ele.me, which is Alibaba's on-demand food delivery platform. IODetector helps Ele.me better understand the courier's working status and improve the delivery efficiency. It is now being deployed in over 300 cities in China and running on the mobile phones of over 1 million couriers. According to real world A/B testing, IODetector helps shorten the average delivery time of each order by 19 seconds and reduce the late delivery rate of orders relatively by 5.6%, which translate to yearly ~108 million RMB cost saving for couriers and restaurants, and ~11 million RMB profit gain for Ele.me platform.
|
|
|
☞ [Project] Urban traffic informatics with crowdsensing (2011-2018)
This traffic sensing system solely relies on the collaborative effort of participating users and is independent from the commercial transportation companies. The real time bus information is collected from the bus riders to predict bus arrival time and estimate the traffic map of the entire urban area. Specifically, we make use of the audio beep signal from card readers to detect the boarding and cellular fingerprint to identify the bus route and track the bus. We implemented a pilot system in Singapore in 2011 and also did evaluation experiments in London in 2012. We further designed and implemented GPTE on top of the Spark, an emerging cluster computing framework. GPTE benefits from its non-linear traffic correlation modeling and the graph-parallel processing framework built on clustered machines. By representing the road network as a property graph, GPTE decomposes the numerous computations involved in non-linear models to vertices and performs traffic estimation via neural network modeling and iterative information propagation. Extensive experiments were performed with real-world data input from Singapore's transport authority and results showed that GPTE achieves as high as 88% accuracy in traffic estimation.
|
|
Honors and Awards
Patents and Technology Disclosure Records
Professional Activities