More Like this

1. Improving Monitoring of Participatory Civil Issue Requests through Optimal Online Classification

   Zois, Daphney-Stavroula; Yong, Christopher; Chelmis, Charalampos; Kapodistria, Angeliki; Lee, Wonhyung (October 2018, Asilomar Conference on Signals, Systems, and Computers)

   Participatory civil issue monitoring has emerged as an easy way for concerned citizens to report problems to their local government. For reported issues to be timely processed and addressed however, accurate, online and real–time processing methods to infer issue types are necessary. To address this challenge, we propose a computational, near–real–time civil issue reports processing method to estimate the actual issue from ambiguous and/or complementary information accurately and efficiently. We demonstrate the effectiveness of the proposed approach using a real-world dataset from SeeClickFix. We show that our approach is both highly accurate and scalable. 

   more » « less
2. Automated Optimal Online Civil Issue Classification using Multiple Feature Sets

   https://doi.org/10.1109/IEEECONF44664.2019.9048766  

   Nazar, Imara; Liyanage, Yasitha Warahena; Zois, Daphney-Stavroula; Chelmis, Charalampos (November 2019, 2019 53rd Asilomar Conference on Signals, Systems, and Computers)

   In this paper, the automatic classification of non-emergency civil issues in crowdsourcing systems is addressed in the case where multiple feature sets are available. We recognize that multiple feature sets can contain useful complementary information regarding the type of an issue leading to a more accurate decision. However, using all features in these sets may delay the decision. Since we are interested in reaching an accurate decision in a timely manner, an optimal way of selecting features from multiple feature sets is needed. To this end, we propose a novel approach that sequentially reviews available features and feature sets to decide whether the feature review process must continue in the current set or move to the next one. In the end, when all feature sets have been reviewed, the issue is classified using all available information. It is shown that the proposed approach is guaranteed to review the least number of features in all feature sets before reaching a decision, while the optimum decision rule is shown to minimize the average Bayes risk. Evaluation on real world SeeClickFix data demonstrates the ability to classify issues by reviewing 99.5% less features than state-of-the-art without sacrificing accuracy. 

   more » « less
3. Automating the Classification of Urban Issue Reports: an Optimal Stopping Approach

   https://doi.org/10.1109/ICASSP.2019.8682778  

   Liyanage, Yasitha Warahena; Zois, Daphney-Stavroula; Chelmis, Charalampos; Yao, Mengfan (May 2019, 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP))

   Empowering citizens to interact directly with their local governments through civic engagement platforms has emerged as an easy way to resolve urban issues. However, for authorities to manually process reported issues is both impractical and inefficient; accurate, online and near-real-time processing methods are necessary to maintain citizens' satisfaction with their local governments. Herein, an optimal stopping framework is proposed to process urban issue requests quickly and accurately. The optimal classification and stopping rules are derived, and significant reduction in time-to-decision without sacrificing accuracy is demonstrated on a real-world dataset from SeeClickFix. 

   more » « less
4. WHAT MATTERS THE MOST? OPTIMAL QUICK CLASSIFICATION OF URBAN ISSUE REPORTS BY IMPORTANCE

   https://doi.org/10.1109/GlobalSIP.2018.8646639  

   Liyanage, Yasitha; Yao, Mengfan; Yong, Christopher; Zois, Daphney-Stavroula; Chelmis, Charalampos (November 2018, 2018 IEEE Global Conference on Signal and Information Processing (GlobalSIP))

   Civic engagement platforms such as SeeClickFix and FixMyStreet have revolutionized the way citizens interact with local governments to report and resolve urban issues. However, recognizing which urban issues are important to the community in an accurate and timely manner is essential for authorities to prioritize important issues, allocate resources and maintain citizens' satisfaction with local governments. To this end, a novel formulation based on optimal stopping theory is devised to infer urban issues importance from ambiguous textual, time and location information. The goal is to optimize recognition accuracy, while minimizing the time to reach a decision. The optimal classification and stopping rules are derived. Furthermore, a near-real-time urban issue reports processing method to infer the importance of incoming issues is proposed. The effectiveness of the proposed method is illustrated on a real-word dataset from SeeClick-Fix, where significant reduction in time-to-decision without sacrificing accuracy is observed. 

   more » « less
5. Penetrating radar combined with 3-D imaging for real-time augmented reality sensing and classification

   https://doi.org/10.1117/12.2587703  

   Girard, Joshua H.; Burns, Dylan C.; Huston, Dryver R.; Xia, Tian (April 2021, SPIE Defense + Commercial Sensing: Virtual, Augmented, and Mixed Reality (XR) Technology for Multi-Domain Operations II)

   Dennison, Mark S.; Krum, David M.; Sanders-Reed, John; Arthur, Jarvis (Ed.)

   This paper presents research on the use of penetrating radar combined with 3-D computer vision for real-time augmented reality enabled target sensing. Small scale radar systems face the issue that positioning systems are inaccurate, non-portable or challenged by poor GPS signals. The addition of modern computer vision to current cutting-edge penetrating radar technology expands the common 2-D imaging plane to 6 degrees of freedom. Applying the fact that the radar scan itself is a vector with length equivalent to depth from the transmitting and receiving antennae, these technologies used in conjunction can generate an accurate 3-D model of the internal structure of any material for which radar can penetrate. The same computer vision device that localizes the radar data can also be used as the basis for an augmented reality system. Augmented reality radar technology has applications in threat detection (human through-wall, IED, landmine) as well as civil (wall and door structure, buried item detection). For this project, the goal is to create a data registration pipeline and display the radar scan data visually in a 3-D environment using localization from a computer vision tracking device. Processed radar traces are overlayed in real time to an augmented reality screen where the user can view the radar signal intensity to identify and classify targets. 

   more » « less