ljnlonoljpiljm/LVIS

该数据集是一个对象检测数据集，包含图像数据以及对应的标注信息。每个样本包括一个唯一的id、图像文件、图像的高度和宽度，以及objects序列。objects序列中每个对象都有边界框（bboxes）和类别标签（classes），类别标签覆盖了从0到695的多种物体类型，如气雾罐、空调、飞机、闹钟、酒精、鳄鱼、杏仁、救护车等，总计696个类别。这些类别涵盖了日常用品、食物、动物、工具、车辆、服装等多种领域，适用于训练和评估计算机视觉模型，特别是对象检测任务。This dataset is an object detection dataset containing image data along with corresponding annotations. Each sample includes a unique id, an image file, the height and width of the image, and an objects sequence. The objects sequence consists of bounding boxes (bboxes) and class labels (classes) for each object, with class labels ranging from 0 to 695, covering a wide variety of object types such as aerosol can, air conditioner, airplane, alarm clock, alcohol, alligator, almond, ambulance, and more, totaling 696 categories. These categories encompass everyday items, food, animals, tools, vehicles, clothing, and other domains, making it suitable for training and evaluating computer vision models, particularly for object detection tasks.