Animal Detection and Classification from Camera Trap Images Using Different Mainstream Object Detection Architectures.

Camera traps are widely used in wildlife surveys and biodiversity monitoring. Depending on its triggering mechanism, a large number of images or videos are sometimes accumulated. Some literature has proposed the application of deep learning techniques to automatically identify wildlife in camera trap imagery, which can significantly reduce manual work and speed up analysis processes. However, there are few studies validating and comparing the applicability of different models for object detection in real field monitoring scenarios. In this study, we firstly constructed a wildlife image dataset of the Northeast Tiger and Leopard National Park (NTLNP dataset). Furthermore, we evaluated the recognition performance of three currently mainstream object detection architectures and compared the performance of training models on day and night data separately versus together. In this experiment, we selected YOLOv5 series models (anchor-based one-stage), Cascade R-CNN under feature extractor HRNet32 (anchor-based two-stage), and FCOS under feature extractors ResNet50 and ResNet101 (anchor-free one-stage). The experimental results showed that performance of the object detection models of the day-night joint training is satisfying. Specifically, the average result of our models was 0.98 mAP (mean average precision) in the animal image detection and 88% accuracy in the animal video classification. One-stage YOLOv5m achieved the best recognition accuracy. With the help of AI technology, ecologists can extract information from masses of imagery potentially quickly and efficiently, saving much time.

Angry characters and frightened souls: Patients and family explanatory models of bipolar disorder in Taiwan.

Bipolar disorder (BD) affects a significant proportion of Taiwanese individuals (Weissman et al., 1996; Yang, Yeh, & Hwu, 2012). Psychotropic medications are typically the mainstay of treatment for BD, and there is an abundance of international research on biological etiology and medication options. However, there is comparatively little research on psychosocial aspects of BD, including how it is understood and managed within families. As culture provides the context in which psychiatric disease is managed, there is a need to identify distinct Chinese psychosocial perspectives that might shed light on intervention options. This research explored how Taiwanese patients and family members comprehend and cope with BD. A sample of 42 participants, including 20 Taiwanese patients diagnosed with Bipolar Disorder-I (BD-I) for at least 4 years, and 22 family members, participated in separate interviews on explanatory models of illness. Qualitative thematic analysis focused on features that were distinct from those in current Western research literature. Five themes were identified that represented Taiwanese conceptualizations of BD, notions of etiology, views regarding treatment, and the difficulties in managing the disorder. Participants used Chinese language terms and descriptions of BD that reflected greater concerns about irritability, anger, and family conflict than about other symptoms, and participants also emphasized characterological trait descriptions of the condition. Their responses reflected their acceptance of lifelong family responsibility for caretaking, clashing beliefs regarding biomedical versus traditional Chinese medical and spiritual models of etiology and cure, profound concerns about the effects of psychiatric medication on the liver and kidney systems, and a focus on stress rather than genetic or biological models of etiology.