ModeNet and NavNet Dataset

该数据集包含了1500张环境摄像头图像，分别用于训练ModeNet和NavNet，以进行模式分类和航点预测。根据与物体的距离阈值和相对航点，定义了地面真实标签。在数据处理中，采用了诸如旋转和对比度调整等数据增强技术。ModeNet达到了0.89的准确率，而NavNet则实现了0.93的准确率。该数据集的规模分为两部分：1500张图像用于模拟环境，6200张图像用于真实世界实验。任务重点在于为机器人任务进行模式分类和航点预测。

This dataset contains 1500 environmental camera images, which are respectively used for training ModeNet and NavNet to perform pattern classification and waypoint prediction. Ground truth labels are defined based on distance thresholds to objects and relative waypoints. Data augmentation techniques such as rotation and contrast adjustment were applied during data processing. ModeNet achieved an accuracy of 0.89, while NavNet reached an accuracy of 0.93. This dataset is divided into two parts in terms of scale: 1500 images are used for simulated environments, and 6200 images are used for real-world experiments. The task focuses on pattern classification and waypoint prediction for robotic tasks.