MBW Zoo Data

We release a challenging dataset consisting image frames of tail-end distribution categories (such as Fish, Colobus Monkeys, Chimpanzees, etc.) with their corresponding 2D, 3D, and Bounding-Box labels generated from minimal human intervention. Some of the prominent use cases of this dataset include not only sparse 2D and 3D landmark prediction, but also dense reconstruction tasks such as dense deformable shape reconstruction, novel view rendering (NeRF), Detection and Tracking, and finally this dataset could also be used for camera estimation in Simultaneous Localization and Mapping (SLAM) frameworks.本团队发布了一项极具挑战性的数据集，该数据集包含长尾分布尾部类别（tail-end distribution categories）的图像帧，例如鱼类、疣猴、黑猩猩等，并附带通过最小人工干预生成的对应2D、3D标注及边界框（Bounding-Box）标注。该数据集的典型应用场景不仅涵盖稀疏2D与3D关键点预测，还包含各类密集重建任务，如密集可变形形状重建、新视角渲染（神经辐射场，NeRF）、目标检测与跟踪；此外，该数据集还可用于同时定位与建图（SLAM）框架中的相机位姿估计。

搜集汇总

背景与挑战

背景概述

MBW Zoo Data是一个包含尾部分布类别（如鱼、疣猴、黑猩猩等）图像帧的数据集，自动生成对应的2D、3D和边界框标签。该数据集适用于稀疏2D/3D关键点预测、密集变形形状重建、新视角渲染（NeRF）、检测跟踪以及SLAM等多种计算机视觉任务。