The z-test based on the Kappa statistic is commonly used to infer superiority of one map production method over another. Typically the same reference data set is used to calculate and next compare the Kappa's of the two maps. This data structure easily leads to dependence between the two error-matrices. This may result in overly large variance estimates and too conservative inference about the difference in accuracy between the two methods. Tests considering the dependency between the error matrices would be more sensitive in such case. In this article we compare the performance of two such tests, a randomization and McNemar's test, with the traditional z-test. We compared 16 alternative methods to classify salt marsh vegetation in The Netherlands. The error matrices were positively associated in all 120 possible comparisons of pairs of classification methods. This suggests that dependency between pairs of error matrices used in classifier comparison is a common phenomenon. Both the randomization and McNemar test gave lower p values and rejected the null hypothesis of equal performance more frequently than the z-test. We therefore recommend considering their use.
To achieve accurate segmentation of thyroid nodule ultrasound images, obtain information on the physiological parameters of the lesion area and guide the clinical formulation of individualized treatment plan, an improved network based on the U2‐Net model is proposed in this paper. Thyroid images of 264 patients and 215 healthy volunteers at the First Hospital of Shanxi Medical University from February 2016 to June 2022 are studied, and the digital database thyroid image (DDTI) data set is proposed for data expansion. The experimental results show that the Dice coefficient on the test set was 80.58%, and the mean intersection over union was 81.21%. The improved U2‐Net model has the best segmentation accuracy compared with similar models, realizes the automatic segmentation of thyroid nodules, provides help for manual segmentation, and has good application prospects and clinical value.
Human detection is the first step for a number of applications such as smart video surveillance, driving assistance systems, and intelligent digital content management. It’s a challenging problem due to the variance of illumination, color, scale, pose, and so forth. This chapter reviews various aspects of human detection in static images and focuses on learning-based methods that build classifiers using training samples. There are usually three modules for these methods: feature extraction, classifier design, and merge of overlapping detections. The chapter reviews most existing methods for each module and analyzes their respective pros and cons. The contribution includes two aspects: first, the performance of existing feature sets on human detection are compared; second, a fast human detection system based on histogram of oriented gradients features and cascaded AdaBoost classifier is proposed. This chapter should be useful for both algorithm researchers and system designers in the computer vision and pattern recognition community.
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.