| Summary: | In response to the challenge of collecting behavioral data on <i>Amur tigers</i> living in forests, a remote real-time data collection approach is proposed. In this article, a novel attention mechanism named CBAM-E is introduced, and CBAM-E as well as the CIoU loss function are incorporated into the YOLOX object detection algorithm, resulting in a new YOLOX model. The new model demonstrates significant performance improvements over the original model, with the <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><msub><mrow><mi>m</mi><mi>A</mi><mi>P</mi></mrow><mrow><mn>0.5</mn></mrow></msub></mrow></semantics></math></inline-formula> detection accuracy metric rising from 97.32 to 98.18%, indicating a boost of 0.86%, and the <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><msub><mrow><mi>m</mi><mi>A</mi><mi>P</mi></mrow><mrow><mn>0.75</mn></mrow></msub></mrow></semantics></math></inline-formula> metric increasing from 75.10 to 78.70%, marking an enhancement of 3.60%. The enhanced algorithm is subsequently applied to remote terminal information collection, offering a reference for detection algorithms in the study of wild behaviors of <i>Amur tigers</i> in forests, biodiversity conservation, and the collection of related field data about <i>Amur tigers</i> in the wild.
|
|---|