vapaau/3D-ZeF

3D-ZeF是第一个RGB 3D多目标跟踪数据集，专门用于斑马鱼的行为分析。斑马鱼是研究神经系统疾病、社交焦虑等广泛使用的模型生物。数据集通过两个GoPro相机、一个玻璃水族箱和两个IKEA灯板的低成本设备捕获，旨在使小型实验室和学生等更容易进行自动化3D行为分析。数据集包含8个序列，每个序列包含两个GoPro的录制视频、地面真实标签和校准文件。所有鱼在所有帧中都进行了手动边界框和点标注。点标注位于鱼的头部，以便在立体3D重建阶段有一个稳定的点。数据集分为训练、验证和测试集，每个集包含多个序列，且每个序列包含两个录制视频。训练和验证集的鱼来自同一批，而测试集的鱼来自另一批更年轻的鱼，以避免数据泄漏和过拟合。

This is the dataset page for 3D-ZeF, the first RGB 3D multiple object tracking dataset of its kind. Zebrafish is a widely used model organism for studying neurological disorders, social anxiety, and more. Behavioral analysis can be a critical part of such research and it has traditionally been conducted manually, which is an expensive, time consuming, and subjective task. The dataset is captured using an off-the-shelf setup consisting of two GoPro cameras, a glass aquarium, and two IKEA light panels, aiming to make automated 3D behavioral analysis of zebrafish more accessible to smaller labs, students, and the likes. The dataset consists of 8 sequences, each containing two recordings from the two GoPros, along with ground truth labels and calibration files. All fish are manually bounding and point annotated in all frames, with point annotations placed on the head for stable stereo 3D reconstruction. The dataset is split into training, validation, and test sets, each containing multiple sequences with two recordings each. Unique fish are used in each split, with training and validation fish from one cohort and test fish from a different, younger cohort to avoid data leakage and overfitting.