Han Liu
· Orrington Lunt Professor of Computer Science and (by courtesy) Industrial Engineering & Management SciencesNorthwestern University · Chemical Engineering
Active 1998–2024
About
Han Liu is the Orrington Lunt Professor of Computer Science at Northwestern University, where he also holds courtesy appointments in Industrial Engineering & Management Sciences. He directs the MAGICS (Modern Artificial General Intelligible and Computer Systems) lab and is the director of the Center for Foundation Models and Generative AI (CFMG) at Northwestern. His research lies at the intersection of artificial intelligence and computer systems, focusing on deploying statistical machine learning methods on edges and clouds to achieve analytical advantages. Liu has previously served as the director of the deep reinforcement learning center at Tencent AI Lab and has held faculty positions at Princeton University and Johns Hopkins University. He received a joint PhD in Machine Learning and Statistics from Carnegie Mellon University, advised by John Lafferty and Larry Wasserman. His work integrates modern AI and computer systems, exploiting large foundation models and probabilistic graphical models to revolutionize science, engineering, and business. Liu has received numerous awards, including the Presidential Early Career Award for Scientists and Engineers, the Alfred P Sloan Fellowship in Mathematics, the IMS Tweedie New Researcher Award, the ASA Noether Young Scholar Award, the NSF CAREER Award, the Howard B Wentz Award, and the Umesh Gavaskar Memorial Best Dissertation Award. He serves as an associate editor for several prominent journals and has been involved as an area chair for major conferences such as NeurIPS, ICML, and ICLR.
Research topics
- Artificial Intelligence
- Computer Science
- Computer vision
- Remote sensing
- Algorithm
- Materials science
- Engineering
- Geography
- Real-time computing
Selected publications
IEEE Sensors Journal · 2021 · 120 citations
- Artificial Intelligence
- Computer Science
- Artificial Intelligence
When a sensor data-based detection method is used to detect the potential defects of industrial products, the data are normally imbalanced. This problem affects improvement of the robustness and accuracy of the defect detection system. In this work, welding defect detection is taken as an example: based on imbalanced radiographic images, a welding defect detection method using generative adversarial network combined with transfer learning is proposed to solve the data imbalance and improve the accuracy of defect detection. First, a new model named contrast enhancement conditional generative adversarial network is proposed, which is creatively used as a global resampling method for data augmentation of X-ray images. While solving the limitation of feature extraction due to low contrast in some images, the data distribution in the images is balanced, and the number of the image samples is expanded. Then, the Xception model is introduced as a feature extractor in the target network for transfer learning, and based on the obtained balanced data, fine-tuning is performed through frozen-unfrozen training to build the intelligent defect detection model. Finally, the defect detection model is used to detect five types of welding defects, including crack, lack of fusion, lack of penetration, porosity, and slag inclusion; an F1-score of 0.909 and defect recognition accuracy of 92.5% are achieved. The experimental results verify the effectiveness and superiority of the proposed defect detection method compared to conventional methods. For other similar applications to defect detection, the proposed method has promotional value.
- RETRACTED
PLoS ONE · 2021 · 409 citations
- Computer Science
- Artificial Intelligence
- Computer Science
This exploration primarily aims to jointly apply the local FCN (fully convolution neural network) and YOLO-v5 (You Only Look Once-v5) to the detection of small targets in remote sensing images. Firstly, the application effects of R-CNN (Region-Convolutional Neural Network), FRCN (Fast Region-Convolutional Neural Network), and R-FCN (Region-Based-Fully Convolutional Network) in image feature extraction are analyzed after introducing the relevant region proposal network. Secondly, YOLO-v5 algorithm is established on the basis of YOLO algorithm. Besides, the multi-scale anchor mechanism of Faster R-CNN is utilized to improve the detection ability of YOLO-v5 algorithm for small targets in the image in the process of image detection, and realize the high adaptability of YOLO-v5 algorithm to different sizes of images. Finally, the proposed detection method YOLO-v5 algorithm + R-FCN is compared with other algorithms in NWPU VHR-10 data set and Vaihingen data set. The experimental results show that the YOLO-v5 + R-FCN detection method has the optimal detection ability among many algorithms, especially for small targets in remote sensing images such as tennis courts, vehicles, and storage tanks. Moreover, the YOLO-v5 + R-FCN detection method can achieve high recall rates for different types of small targets. Furthermore, due to the deeper network architecture, the YOL v5 + R-FCN detection method has a stronger ability to extract the characteristics of image targets in the detection of remote sensing images. Meanwhile, it can achieve more accurate feature recognition and detection performance for the densely arranged target images in remote sensing images. This research can provide reference for the application of remote sensing technology in China, and promote the application of satellites for target detection tasks in related fields.
A YOLOv3-based Learning Strategy for Real-time UAV-based Forest Fire Detection
2020 · 84 citations
- Computer Science
- Artificial Intelligence
- Computer Science
Forest resources safety is of paramount importance for natural and public security. Forest fire detection methods have been attracted much attention recently, but the performance in terms of comprehensiveness, rapidity, and accuracy is still not satisfactory. A deep learning fire detection algorithm is proposed in this paper, aiming at improving the detection accuracy and efficiency by using the unmanned aerial vehicle (UAV). A large-scale YOLOv3 network is firstly developed which can ensure the detection accuracy. The algorithm is then applied to UAV forest fire detection (UAV-FFD) platform, where the fire images can be captured by the UAV and transmitted to the ground-station in real time. The testing results indicate that the recognition rate of the detection algorithm is about 91%, and the frame rate can reach up to 30 FPS (Frames Per Second). It shows strong potential in real-time application for precision forest fire detection.
Recent grants
NSF · $238k · 2017–2020
NSF · $359k · 2017–2021
NSF · $401k · 2015–2018
NSF · $500k · 2014–2018
CAREER: An Integrated Inferential Framework for Big Data Research and Education
NSF · $348k · 2017–2023
Frequent coauthors
- 30 shared
Ding Liu
Hanzhong University
- 29 shared
Tuo Zhao
Georgia Institute of Technology
- 26 shared
Zhaoran Wang
Shanghai University
- 24 shared
Guo Xie
Hebei University of Engineering
- 24 shared
Youmin Zhang
- 21 shared
Wenqing Wang
Xi'an University of Technology
- 18 shared
Qiang She
Chongqing Medical University
- 17 shared
Tong Zhang
Tongji University
Education
- 2003
Doctor, Control Engineering
Xian University of Technology
Awards & honors
- Alfred P Sloan Fellowship in Mathematics
- IMS Tweedie New Researcher Award
- ASA Noether Young Scholar Award
- NSF CAREER Award
- Howard B Wentz Award
Similar researchers at Northwestern University
- Resume-aware match score
- Save to shortlist
- AI-drafted outreach
See your match with Han Liu
PhdFit ranks faculty by your research interests, methods, and publications — grounded in their actual work, not templates.
- Free to start
- No credit card
- 30-second signup