Common benchmark data sets, standardized performance metrics, and baseline algorithms have demonstrated considerable impact on research and development in a variety of application domains. These resources provide both consumers and developers of technology with a common framework to objectively compare the performance of different algorithms and algorithmic improvements. In this paper, we present such a framework for evaluating object detection and tracking in video: specifically for face, text, and vehicle objects. This framework includes the source video data, ground-truth annotations (along with guidelines for annotation), performance metrics, evaluation protocols, and tools including scoring software and baseline algorithms. For each detection and tracking task and supported domain, we developed a 50-clip training set and a 50-clip test set. Each data clip is approximately 2.5 minutes long and has been completely spatially/temporally annotated at the I-frame level. Each task/domain, therefore, has an associated annotated corpus of approximately 450,000 frames. The scope of such annotation is unprecedented and was designed to begin to support the necessary quantities of data for robust machine learning approaches, as well as a statistically significant comparison of the performance of algorithms. The goal of this work was to systematically address the challenges of object detection and tracking through a common evaluation framework that permits a meaningful objective comparison of techniques, provides the research community with sufficient data for the exploration of automatic modeling techniques, encourages the incorporation of objective evaluation into the development process, and contributes useful lasting resources of a scale and magnitude that will prove to be extremely useful to the computer vision research community for years to come.
Summary1. Quercus macrocarpa (bur oak) is the dominant tree species along much of the prairie-forest border in the northern-central United States, and movement of Q. macrocarpa in response to climate change may determine the rate at which the prairie-forest ecotone shifts. To investigate likely controls over Q. macrocarpa performance at the edge of its range, we used tree rings to establish the links between drought, growth-rate and mortality for three sites spanning the prairie-forest border in Minnesota. 2. Quercus macrocarpa growth during the 20th century correlates strongly with the Palmer Drought Severity Index (PDSI) and more weakly with raw temperature and precipitation values for all three sites. However, the sensitivity of annual growth rates to drought has steadily declined over time as evidenced by increasing growth residuals and higher growth rates for a given PDSI value after 1950 compared with the first half of the century. We hypothesize that increased atmospheric carbon dioxide concentration may lead to increased water-use efficiency, although we cannot rule out other environmental factors. 3. Because growth is an excellent predictor of Q. macrocarpa mortality, growth-climate relationships provide information on whether oak forests will contract, because of individual tree death, when climate changes. For Q. macrocarpa, declining sensitivity of growth to drought translates into lower predicted mortality rates at all sites. At one site, declining moisture sensitivity yields a 49% lower predicted mortality from a severe drought (PDSI = )8, on par with the worst 1930s 'American Dust Bowl' droughts in our study region). 4. Unless the changing relationship between growth and climate is incorporated into forest simulation models, the predicted rate of established tree dieback in a warmer, drier climate may be exaggerated. 5. Synthesis. Adult Quercus macrocarpa trees appear to be increasingly insensitive to droughtinduced mortality. Because the species is dominant at the prairie-forest ecotone in the northern-central United States, movement of the ecotone in response to climate change may be delayed for decades.
No abstract
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.