The TYO-L (Toyota Light) dataset is part of the Benchmark for 6D Object Pose Estimation (BOP). Let's delve into the details:
-
TYO-L Dataset Overview:
- Objects: The TYO-L dataset contains 21 objects.
- Capture Setup: These objects were captured in multiple poses on a table-top setup.
- Variations: The dataset includes four different tablecloths and five different lighting conditions.
- Texture Mapping: The objects are represented by texture-mapped 3D models.
- License: The dataset is licensed under CC BY-NC 4.0¹.
-
Object Characteristics:
- The objects in TYO-L exhibit a wide range of sizes.
- While the dataset focuses on pose estimation, it provides valuable information for research related to low-light image and video enhancement as well ².
-
Benchmark for 6D Object Pose Estimation (BOP):
- BOP aims to advance the field of 6D object pose estimation by providing standardized datasets, evaluation metrics, and challenges.
- Other datasets within BOP include LM (Linemod), LM-O (Linemod-Occluded), and T-LESS.
- Each dataset includes 3D object models, training/test RGB-D images, and annotations for ground-truth 6D object poses, 2D bounding boxes, and 2D binary masks.
- The datasets cover a variety of scenarios, including texture-less objects, occlusion, and symmetrical shapes ¹.
In summary, the TYO-L dataset contributes to the advancement of object pose estimation research, particularly in low-light conditions, and provides valuable resources for the scientific community.
(1) Datasets - BOP: Benchmark for 6D Object Pose Estimation. https://bop.felk.cvut.cz/datasets/.
(2) Low-Light Image and Video Enhancement: A Comprehensive Survey and Beyond. https://arxiv.org/html/2212.10772v5.
(3) Toyota Dataset | Kaggle. https://www.kaggle.com/datasets/chinmaypradhan29/toyota-dataset.
(4) undefined. https://bop.felk.cvut.cz/media/data/bop_datasets.