Plant identification is an interesting and challenging research topic due to the variety of plant species. Among different parts of the plant, leaf is widely used for plant identification because it is usually the most abundant type of data available in botanical reference collections and the easiest to obtain in the field studies. A number of works have been done for plant leaf identification. However, this topic is still an open research topic. Recently, kernel descriptor (KDES) has been proposed and proved to be robust for many object recognition applications. In [1], we have proposed to use this descriptor for leaf-based plant identification. The contributions of this paper are three-fold. Firstly, we present improved kernel descriptor. Secondly, we propose a fully automatic leaf-based plant identification method consisting of petiole detection, plant leaf orientation normalization, and plant leaf identification. Finally, we introduce a new plant identification application on Android devices: Vietnamese medicinal plant search. This application provides both search modalities: text-based and image-based.
We aimed to investigate the performance of a chest X-ray (CXR) scoring scale of lung injury in prediction of death and ICU admission among patients with COVID-19 during the 2021 peak pandemic in HCM City, Vietnam. CXR and clinical data were collected from Vinmec Central Park-hospitalized patients from July to September 2021. Three radiologists independently assessed the day-one CXR score consisting of both severity and extent of lung lesions (maximum score = 24). Among 219 included patients, 28 died and 34 were admitted to the ICU. There was a high consensus for CXR scoring among radiologists (κ = 0.90; CI95%: 0.89–0.92). CXR score was the strongest predictor of mortality (tdAUC 0.85 CI95% 0.69–1) within the first 3 weeks after admission. A multivariate model confirmed a significant effect of an increased CXR score on mortality risk (HR = 1.33, CI95%: 1.10 to 1.62). At a threshold of 16 points, the CXR score allowed for predicting in-hospital mortality and ICU admission with good sensitivity (0.82 (CI95%: 0.78 to 0.87) and 0.86 (CI95%: 0.81 to 0.90)) and specificity (0.89 (CI95%: 0.88 to 0.90) and 0.87 (CI95%: 0.86 to 0.89)), respectively, and can be used to identify high-risk patients in needy countries such as Vietnam.
Gesture recognition has important applications in sign language and human -machine interfaces. In recent years, recognizing dynamic hand gesture using multi-modal data has become an emerging research topic. The problem is challenging due to the complex movements of hands and the limitations of data acquisition. In this work, we present a new approach for recognizing hand gesture using motion history images (MHI) [1] and a kernel descriptor (KDES) [2]. We propose to use an improved version of MHI for modeling movements of hand gesture, where MHI is computed on both RGB and depth data. We propose some improvements in patch-level feature extraction for KDES, which is then applied to MHI to represent gesture features. Then SVM classifier is trained for recognizing gestures. Experiments have been conducted on challenging hand gesture data set of CHALEARN contest [3]. An extensive investigation has been done to analyze the performance of both improved MHI and KDES on multi-modal data. Experimental results show the state-of-the-art of our approach in comparison to the results of the contest.
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.