ZixunH/DTTD3_Impedance

🦾 DTTRD-E: DTTD's Robotics Dataset Extension

If you want to use this dataset in your ongoing / future research, feel free to contact us for details: [email protected]; my collaborator: [email protected]; and my advisor: [email protected]

Download the robotics dataset extension: https://huggingface.co/datasets/ZixunH/DTTD3_Impedance/tree/main

This dataset follows the same folder structure as our previous work, DTTD-Mobile, but is 3x larger and focuses on small, textureless industrial objects with around 52,000 frames. Estimating 6DoF poses is even MORE challenging due to significant occlusion from robotic graspers—yet this reflects the realistic conditions robots face in factory environments.

📁 Folder Structure (including cad models, rgb images, semantic images, depth images, camera intrinsics, object poses, end-effector poses, etc):

Robust-Digital-Twin-Tracking
├── checkpoints
│   ├── m8p4.pth
│   ├── m8p4_filter.pth
│   └── ...
|   ...
└── dataset
    └── dttrd-e
        ├── dataset_config
        ├── dataset.py
        └── DTTD_Robotic_Dataset_Extension
            └── root
                ├── cameras
                │   ├── az_camera1 (checkout DTTDv1)
                │   ├── iphone14pro_camera1 (checkout our DTTD-Mobile)
                │   ├── real_sense (checkout DTTRD-E)
                │   └── ZED2 (TBA...)
                ├── data
                │   ├── scene1
                │   │   └── data
                │   │   │   ├── 00001_color.jpg
                │   │   │   ├── 00001_depth.png
                │   │   │   └── ...
                |   │   └── scene_meta.yaml
                │   ├── scene2
                │   │   └── data
                |   │   └── scene_meta.yaml
                │   ...
                └── objects
                    ├── waterproof
                    │   ├── waterproof.mtl
                    │   ├── waterproof.obj
                    │   ├── front.xyz
                    │   ├── points.xyz
                    │   ├── ...
                    │   └── textured.obj.mtl
                    ├── white_charger
                    └── ...

And if our work is useful or relevant to your research, please kindly recognize our contributions by citing the 2025 CVPRW paper Robust 6DoF Pose Estimation Against Depth Noise and a Comprehensive Evaluation on a Mobile Dataset:

@InProceedings{Huang_2025_CVPR,
    author    = {Huang, Zixun and Yao, Keling and Zhao, Zhihao and Pan, Chuanyu and Yang, Allen},
    title     = {Robust 6DoF Pose Estimation Against Depth Noise and a Comprehensive Evaluation on a Mobile Dataset},
    booktitle = {Proceedings of the Computer Vision and Pattern Recognition Conference (CVPR) Workshops},
    month     = {June},
    year      = {2025},
    pages     = {1848-1857}
}