Hongming Zhu

Hongming Zhu


Research Focus & Metrics

Academic Career Length

13Years (2014-Present)

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

Representative Publications

View All Publications
4 January 2024 in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Verified
Journal Article
Open access
YLYizhang Lin
Yizhang Lin
Sicong LiuSicong Liu
Sicong Liu
YZYongjie Zheng
Yongjie Zheng
XTXiaohua Tong
Xiaohua Tong
HXHuan Xie
Huan Xie
JZJie Zhang
Jie Zhang
Multitemporal change detection (CD) plays a crucial role in the remote sensing application field. In recent years, supervised deep learning methods have shown excellent performance in detecting changes in very-high-resolution (VHR) images. However, these methods require a large number of labeled samples for training, making the process time-consuming and labor-intensive. Unsupervised approaches are more attractive in practical applications since they can produce a CD map without relying on any ground reference or prior knowledge. In this article, we propose a novel unsupervised CD approach, named transformer-based multivariate alteration detection (trans-MAD). It utilizes a pre-detection strategy that combines the compressed change vector analysis and the iteratively reweighted multivariate alteration detection (IR-MAD) to generate reliable pseudotraining samples. More accurate and robust CD results can be achieved by leveraging the IR-MAD to detect insignificant changes and by incorporating the transformer-based attention mechanism to model the difference or similarity between two distant pixels in an image. The proposed trans-MAD approach was validated on two VHR bitemporal satellite remote sensing datasets, and the obtained experimental results demonstrated its superiority comparing with the state-of-the-art unsupervised CD methods.

22 June 2023 in Computers in Biology and Medicine
Verified
Journal Article
Open access
KYKainan Yao
Kainan Yao
XWXiaowen Wang
Xiaowen Wang
WLWannian Li
Wannian Li
HZHongming Zhu
Hongming Zhu
YJYizhi Jiang
Yizhi Jiang
QLQin Liu
Qin Liu
Identification of drug-target interactions (DTIs) is an important step in drug discovery and drug repositioning. In recent years, graph-based methods have attracted great attention and show advantages on predicting potential DTIs. However, these methods face the problem that the known DTIs are very limited and expensive to obtain, which decreases the generalization ability of the methods. Self-supervised contrastive learning is independent of labeled DTIs, which can mitigate the impact of the problem. Therefore, we propose a framework SHGCL-DTI for predicting DTIs, which supplements the classical semi-supervised DTI prediction task with an auxiliary graph contrastive learning module. Specifically, we generate representations for the nodes through the neighbor view and meta-path view, and define positive and negative pairs to maximize the similarity between positive pairs from different views. Subsequently, SHGCL-DTI reconstructs the original heterogeneous network to predict the potential DTIs. The experiments on the public dataset show that SHGCL-DTI has significant improvement in different scenarios, compared with existing state-of-the-art methods. We also demonstrate that the contrastive learning module improves the prediction performance and generalization ability of SHGCL-DTI through ablation study. In addition, we have found several novel predicted DTIs supported by the biological literature. The data and source code are available at: https://github.com/TOJSSE-iData/SHGCL-DTI.

1 June 2023 in Bioinformatics
Verified
Journal Article
Open access
PDF file
XWXiaowen Wang
Xiaowen Wang
HZHongming Zhu
Hongming Zhu
DCDanyi Chen
Danyi Chen
YYYongsheng Yu
Yongsheng Yu
QLQin Liu
Qin Liu
Motivation: Drug combination therapy shows significant advantages over monotherapy in cancer treatment. Since the combinational space is difficult to be traversed experimentally, identifying novel synergistic drug combinations based on computational methods has become a powerful tool for pre-screening. Among them, methods based on deep learning have far outperformed other methods. However, most deep learning-based methods are unstable and will give inconsistent predictions even by simply changing the input order of drugs. In addition, the insufficient experimental data of drug combination screening limits the generalization ability of existing models. These problems prevent the deep learning-based models from being in service. Results: In this article, we propose CGMS to address the above problems. CGMS models a drug combination and a cell line as a heterogeneous complete graph, and generates the whole-graph embedding to characterize their interaction by leveraging the heterogeneous graph attention network. Based on the whole-graph embedding, CGMS can make a stable, order-independent prediction. To enhance the generalization ability of CGMS, we apply the multi-task learning technique to train the model on drug synergy prediction task and drug sensitivity prediction task simultaneously. We compare CGMS’s generalization ability with six state-of-the-art methods on a public dataset, and CGMS significantly outperforms other methods in the leave-drug combination-out scenario, as well as in the leave-cell line-out and leave-drug-out scenarios. We further present the benefit of eliminating the order dependency and the discrimination power of whole-graph embeddings, interpret the rationality of the attention mechanism, and verify the contribution of multi-task learning. Availability and implementation: The code of CGMS is available via https://github.com/TOJSSE-iData/CGMS.

18 March 2020 in Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies
Verified
Journal Article
Open access
MLMan Luo
Man Luo
BDBowen Du
Bowen Du
KKKonstantin Klemmer
Konstantin Klemmer
HZHongming Zhu
Hongming Zhu
HFHakan Ferhatosmanoglu
Hakan Ferhatosmanoglu
HWHongkai Wen
Hongkai Wen
The future of urban mobility is expected to be shared and electric. It is not only a more sustainable paradigm that can reduce emissions, but can also bring societal benefits by offering a more affordable on-demand mobility option to the general public. Many car sharing service providers as well as automobile manufacturers are entering the competition by expanding both their EV fleets and renting/returning station networks, aiming to seize a share of the market and to bring car sharing to the zero emissions level. During their fast expansion, one determinant for success is the ability of predicting the demand of stations as the entire system is growing continuously. There are several challenges in this demand prediction problem: First, unlike most of the existing work which predicts demand only for static systems or at few stages of expansion, in the real world we often need to predict the demand as or even before stations are being deployed or closed, to provide information and decision support. Second, for the new stations to be deployed, there is no historical data available to help the prediction of their demand. Finally, the impact of deploying/closing stations on the other stations in the system can be complex. To address these challenges, we formulate the demand prediction problem in the context of fast expanding electric vehicle sharing systems, and propose a data-driven demand prediction approach which aims to model the expansion dynamics directly from the data. We use a local temporal encoding process to handle the historical data for each existing station, and a dynamic spatial encoding process to take correlations between stations into account with Graph Convolutional Neural Networks (GCN). The encoded features are fed to a multi-scale predictor, which forecasts both the long-term expected demand of the stations and their instant demand in the near future. We evaluate the proposed approach with real-world data collected from a major EV sharing platform for one year. Experimental results demonstrate that our approach significantly outperforms the state of the art, showing up to three-fold performance gain in predicting demand for the expanding EV sharing systems.

21 September 2017
Verified
Book Chapter
Open access
PDF file
XFXavier Ferre
Xavier Ferre
Elena Villalba MoraElena Villalba Mora
Elena Villalba Mora
HJHéctor Julio
Héctor Julio
HZHongming Zhu
Hongming Zhu
Mobile application development is characterized by reduced development cycles and high time-to-market pressure. Usability evaluation in mobile applications calls for the application of cost-effective methods, specially adapted to such constraints. We propose extending the Google Analytics for Mobile Applications basic service to store specific low-level user actions of interest for usability evaluation purposes. The solution can serve both for lab usability testing, automating quantitative data gathering, and for logging real use after application release. It is based on identification of relevant user tasks and the detailed events worth gathering, instrumentation of specific code for data gathering, and subsequent data extraction for calculating relevant usability–related variables. We validated our application in a real usability test by comparing the automatically gathered data with the information gathered by the human observer. Results shows both measurements are statistically exchangeable, opening promising new ways to perform usability testing cost-effectively and at greater scale.

18 May 2021 in Remote Sensing
Verified
Journal Article
Open access
PDF file
QLQin Liu
Qin Liu
LHLetong Han
Letong Han
RTRui Tan
Rui Tan
HFHongfei Fan
Hongfei Fan
WLWeiqi Li
Weiqi Li
Sicong LiuSicong Liu
Sicong Liu
Pansharpening aims at fusing the rich spectral information of multispectral (MS) images and the spatial details of panchromatic (PAN) images to generate a fused image with both high resolutions. In general, the existing pansharpening methods suffer from the problems of spectral distortion and lack of spatial detail information, which might prevent the accuracy computation for ground object identification. To alleviate these problems, we propose a Hybrid Attention mechanism-based Residual Neural Network (HARNN). In the proposed network, we develop an encoder attention module in the feature extraction part to better utilize the spectral and spatial features of MS and PAN images. Furthermore, the fusion attention module is designed to alleviate spectral distortion and improve contour details of the fused image. A series of ablation and contrast experiments are conducted on GF-1 and GF-2 datasets. The fusion results with less distorted pixels and more spatial details demonstrate that HARNN can implement the pansharpening task effectively, which outperforms the state-of-the-art algorithms.