Datasets and Benchmarks

Robotics

KinArt3D_dataset, RLLAB, 2022
- Paper
  - JaeGoo Choy, Geonho Cha, and Songhwai Oh, "Unsupervised 3D Link Segmentation of Articulated Objects with a Mixture of Coherent Point Drift," IEEE Robotics and Automation Letters, vol. 7, no. 3, pp. 7124-7131, Jul. 2022.
- Features
  - Unlike the existing point cloud datasets, all objects in the proposed dataset are composed of articulated objects.
  - The dataset provides motion-label annotations (not category-level), so it is more suitable for the motion-based segmentation task.

R3-Driving-Dataset, RLLAB, 2022
- Paper
  - Jeongwoo Oh, Gunmin Lee, Jeongeun Park, Wooseok Oh, Jaeseok Heo, Hojun Chung, Do Hyung Kim, Byungkyu Park, Chang-Gun Lee, Sungjoon Choi, and Songhwai Oh, "Towards Defensive Autonomous Driving: Collecting and Probing Driving Demonstrations of Mixed Qualities," in Proc. of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Oct. 2022.
- Features
  - R3 Driving Dataset provides various qualities of driving data, not only expert cases but also abnormal cases.
  - We collected abnormal driving data under satisfying safety with the control tower in the simulation testbed at Future Mobility Technology Center (FMTC).

D4RL , 2020
- Open-source benchmark for offline reinforcement learning
- Provide standardized environments and datasets for training and benchmarking algorithms.

ManiSkill, 2021
- Composed of 4 manipulation tasks and various types of objects are available.
- Demonstrations of tasks are provided in the type of ‘point-cloud’ which can be transformed in ‘RGB-D’ type.

CALVIN, 2022
- Manipulation dataset for multi-task-problem. 34 tasks are available in the same environment.
- Demonstrations of tasks are provided in the various views of RGB-D observation. Language labels of the demonstrations are also provided.

Habitat Matterport Dataset, 2021
- The Habitat-Matterport 3D Research Dataset (HM3D) is the largest-ever dataset of 3D indoor spaces.
- It consists of 1,000 high-resolution 3D scans (or digital twins) of building-scale residential, commercial, and civic spaces generated from real-world environments.

Replica Dataset, 2019
- Dataset of high quality reconstructions of a variety of indoor spaces.
- Each reconstruction has clean dense geometry, high resolution and high dynamic range textures, glass and mirror surface information, planar segmentation as well as semantic class and instance segmentation.

3D Scene Graph, 2019
- A Structure for Unified Semantics, 3D Geometry and Camera.
- Given a 3D mesh and registered panoramic images, we construct a graph that spans the entire building and includes semantics on objects (e.g., class, material, and other attributes), rooms (e.g., scene category, volume, etc.) and cameras (e.g., location, etc.), as well as the relationships among these entities.

Gibson Env, 2018
- Real-World Perception for Embodied Agents. Perceptual and physics Simulator.
- Gibson is a virtual environment based off of real-world, as opposed to games or artificial environments, to support learning perception. Gibson enables developing algorithms that explore both perception and action hand in hand.

Matterport3D, 2017
- Large-scale RGB-D dataset containing 10,800 panoramic views from 194,400 RGB-D images of 90 building-scale scenes.

CIFAR-10, 2009
- The CIFAR-10 dataset consists of 60000 32x32 colour images in 10 classes, with 6000 images per class. There are 50000 training images and 10000 test images.