Orienting Novel 3D Objects Using Self-Supervised Learning of Rotation Transforms
Open accessAbstract
Orienting objects is a critical component in the automation of many packing and assembly tasks. We present an algorithm to orient novel objects given a depth image of the object in its current and desired orientation. We formulate a self-supervised objective for this problem and train a deep neural network to estimate the 3D rotation as parameterized by a quaternion, between these current and desired depth images. We then use the trained network in a proportional controller to re-orient objects based on the estimated rotation between the two depth images. Results suggest that in simulation we can rotate unseen objects with unknown geometries by up to 30° with a median angle error of 1.47° over 100 random initial/desired orientations each for 22 novel objects. Experiments on physical objects suggest that the controller can achieve a median angle error of 4.2° over 10 random initial/desired orientations each for 5 objects.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Devgon et al. developed a self-supervised learning model to estimate an object's rotation between desired and current rotation using two depth images.
“Devgon et al. [21] developed a self-supervised learning model to estimate an object's rotation between the desired rotation and the current rotation. They used a trained model to estimate the rotation between two depth images.”
From Self-supervised learning: A succinct review (2023) · cites this paper as [21]
The citation lists haven’t been fetched yet
References 35 · Cited by 11
- Authors
- Devgon, S., Ichnowski, J., Balakrishna, A., Zhang, H., & Goldberg, K.
- Affiliations
- University of California, Berkeley (US)
- Venue
- 2020 IEEE 16th International Conference on Automation Science and Engineering (CASE)
- Publisher
- IEEE
- Published
- 2020-08
- Pages
- 1453-1460
- ISSN
- 2161-8070 (ISSN-L)
- OpenAlex
- W3092152581
- Cited by
- 11
- References
- 35
- Citations by year
- 2026: 1 · 2025: 1 · 2024: 3 · 2023: 1 · 2022: 3 · 2021: 2
- FWCI
- 2.24
- Type
- conference paper
- Language
- en
- Open access
- green
- arxiv.org
- Licence
- Locations
- IEEE International Conference on Automation Science and Engineering (CASE) (publishedVersion)
- arXiv (Cornell University) (submittedVersion, cc-by)
- arXiv (Cornell University) (cc-by)
- Funders
- National Science Foundation (1734633)
- Nvidia
- Toyota Research Institute
- Topics
- Physical Sciences › Engineering › Control and Systems Engineering › Robot Manipulation and Learning
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Physical Sciences › Computer Science › Artificial Intelligence › Reinforcement Learning in Robotics
- Keywords
- object orientation, rotation estimation, quaternions, depth images, self-supervised learning, proportional control, robotic manipulation, unseen object generalization
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
