jomoll/silent-heart-dataset-p08

该数据集包含了图片和对应的文字描述，图片可能包含心形图案或其他形状。数据集中的字段包括图片ID、图片本身、文字描述、描述ID、图片高度H（包括心形decoy和其他类别heart）、背景类型、心率(HR)、心率(LR)、非目标形状(decoy_shape)、命名物体(named_objects)、静默物体(silent_objects)、特殊物体(special_object)、是否提及心形(mention_heart)以及目标概率(p_target)。数据集分为训练集、验证集和测试集，分别包含400,000、4,000和4,000个样本。

The dataset consists of images and their corresponding captions. The images may contain heart shapes or other shapes. The fields in the dataset include image ID, the image itself, caption, caption ID, image height H (including decoy for heart shapes and other heart categories), background type, heart rate (HR), lung rate (LR), non-target shape (decoy_shape), named objects (named_objects), silent objects (silent_objects), special object (special_object), mention of heart (mention_heart), and target probability (p_target). The dataset is split into training, validation, and test sets, containing 400,000, 4,000, and 4,000 samples respectively.