Taiwan-black-bear-yolo
收藏资源简介:
台湾黑熊侦测数据集是一个专门用于训练台湾黑熊(Ursus thibetanus formosanus)物体检测模型的计算机视觉数据集。数据集包含18,112张图像,其中9,429张包含黑熊,11,099个黑熊实例标注,采用YOLO格式,兼容YOLOv5/v8等主流框架。数据集分为训练集(15,013张图像)和验证集(3,099张图像),并包含8,683张背景图像(山羌、水鹿等)以减少误判。数据来源包括新闻影片截图、纪录片帧和公开影像,经过多种增强处理(如仿夜视、去噪等)。数据集适用于野生动物保育、濒危物种监测等非商业用途,采用CC BY-NC 4.0许可。
The Formosan Black Bear Detection Dataset is a computer vision dataset specifically designed for training object detection models for Formosan black bears (Ursus thibetanus formosanus). The dataset consists of 18,112 images, among which 9,429 images contain black bears, with a total of 11,099 annotated bear instances. It follows the YOLO annotation format and is compatible with mainstream frameworks such as YOLOv5 and YOLOv8. The dataset is split into a training set (15,013 images) and a validation set (3,099 images), and also includes 8,683 background images (such as Formosan Reeve's muntjac and Sambar deer) to reduce false positives. The data sources include news video screenshots, documentary frames, and public footage, which have undergone various data augmentation operations such as night vision simulation and denoising. This dataset is intended for non-commercial use cases such as wildlife conservation and endangered species monitoring, and is licensed under CC BY-NC 4.0.



