Leaderboard

You are here: Home » Community » Technical Committees » Data Fusion » Data Fusion Contest Data Fusion Contest Hyperspectral and LiDAR fusion The Data Fusion Contest is organized by the Data Fusion Technical Committee of the IEEE Geoscience and Remote Sensing Society (GRSS). The Committee serves as a global, multi-disciplinary, network for geospatial data fusion, with the objective of connecting people and resources, educating students and professionals, and promoting the best practices in data fusion applications. The Contest is open not only to IEEE members, but to everyone, with the goal of evaluating existing methodologies at the research or operational level to solve remote sensing problems using data from a variety of sensors. This year, the Contest involves two datasets – a hyperspectral image and a LiDAR derived Digital Surface Model (DSM), both at the same spatial resolution (2.5m). The hyperspectral imagery has 144 spectral bands in the 380 nm to 1050 nm region. The dataset was acquired over the University of Houston campus and the neighboring urban area. The Contest consists of two parallel competitions. Users are welcome to participate in one or both of them. Best Paper Award, with the objective of promoting novel synergetic use of hyperspectral and LiDAR data. The deliverable will be a 4-page IEEE-style manuscript that addresses the problem, methodology, results and discussion. We encourage the participants to consider various open problems in the realm of multi-sensor data fusion and to use the dataset provided to demonstrate novel and effective approaches to addressing these problems. Best Classification Award, to promote innovation in classification algorithms, and to provide objective and fair performance comparisons among state-of-the-art algorithms. For this task, users will be provided with ground truth to train and gauge the efficacy of their algorithms. Participants will use the ground truth provided to them, along with the multi-sensor dataset, and submit their classification maps and a brief description of the algorithm. Important Dates Best Classification Award – Results to be submitted between February 16, 2013 and May 1, 2013. Best Paper Award – Manuscripts to be submitted by May 31, 2013 Discussion Forum Questions, technical or otherwise, should be submitted only to the IEEE GRSS Data Fusion Discussion Forum at http://www.linkedin.com/groups/IEEE-GRSS-Data-Fusion-Discussion-367843 How to Obtain the Data and Enter the Contest The data can be requested at http://hyperspectral.ee.uh.edu/?page_id=459 Participants must read and accept the Contest Terms and Conditions. This year, the Contest involves two datasets – a hyperspectral image and a LiDAR derived DSM, both at the same spatial resolution (2.5m). The hyperspectral imagery consists of 144 spectral bands in the 380 nm to 1050 nm region and has been calibrated to at-sensor spectral radiance units. The corresponding co-registered DSM consists of elevation in meters above sea level (per the Geoid 2012A model). The data was acquired by the NSF-funded Center for Airborne Laser Mapping (NCALM) over the University of Houston campus and the neighboring urban area. To enter the Best Paper Award challenge, participants are required to submit a manuscript that will be judged in terms of sound scientific reasoning, problem definition, methodology, validation, and presentation. The deadline for the submission is May 31, 2013. Reports must be 4-page, double column, single spaced, and formatted in accordance with the IEEE International Geoscience and Remote Sensing Symposium template (available at http://igarss2013.org/Papers/PaperKit.html#PartIIIa). Submissions should clearly present the theoretical rationale, experimental details, results, and discussions. The document must be in English and submitted in PDF format. Make sure to include the following: title first and last names of each of the authors affiliations contact email Submit your manuscript electronically to [email protected] To enter the Best Classification Award challenge, participants are required to submit a classification map employing the training samples provided along with the multi-sensor datasets. The validation samples that the Contest organizers will use to evaluate the submissions will not be disclosed. Each team will be allowed to submit one (and only one) classification result between February 16 and May 1, 2013. The organizers will use measurements of statistical significance among accuracies obtained from the submitted maps, to select the winning team. If there is a tie in submissions, the teams in a tie will be asked to submit a manuscript (according to the Best Paper Award requirements defined above) by May 31, 2013. The Award Committee will evaluate these contributions to select the winning team of the Best Classification Award category. The training samples and instructions on how to submit the results will be provided by February 10, 2013