ConsInv Dataset
Introduced by Bojko et al. in Self-Improving SLAM in Dynamic Environments: Learning When to Mask
ConsInv is a stereo RGB + IMU dataset designed for Dynamic SLAM testing and contains two subsets:
- ConsInv-Indoors contains sequences in an office setting where small objects are moved.
- ConsInv-Outdoors contains sequences in an urban environment, where cars and/or people move.
The novelty of ConsInv dataset is 1) the controlled degree of difficulty, from easy to very hard, and 2) the fact that the difficulty of the sequences comes only from object motion: relative motion between camera and object, motion ambiguity, challenging points of view when objects move. The difficulty does not come from motion speed, lack of features, lens flare, etc. - typically seen in other SLAM datasets.
Papers
| Paper | Code | Results | Date | Stars |
|---|
Dataset Loaders
No data loaders found. You can
submit your data loader here.
No data loaders found. You can
submit your data loader here.
