Clustering for pose estimation #258
Assignees
Labels
enhancement
New feature or request
Comments
|
Do we know any paper/source for the implementation of this by AMCL? |
|
Not that I know of. That said, the idea of clustering has been floating around for a long while e.g. https://arxiv.org/pdf/cs/0204044.pdf. |
nahueespinosa
added a commit
that referenced
this issue
Jul 15, 2024
### Proposed changes Addresses #258 and #172. Add new cluster based estimator. - Splits the space in a grid and determines the accumulated weight in each cell. - Starting from the weight peaks, it groups cells into clusters of ever decreasing weight. It creates as many clusters as needed to cover all the particle occupied space. - Estimates mean and covariance for each. - Returns the mean and covariance of the cluster with the highest cumulative weight. #### Type of change - [ ] 🐛 Bugfix (change which fixes an issue) - [x] 🚀 Feature (change which adds functionality) - [ ] 📚 Documentation (change which fixes or extends documentation) ### Checklist - [x] Lint and unit tests (if any) pass locally with my changes - [x] I have added tests that prove my fix is effective or that my feature works - [x] I have added necessary documentation (if appropriate) - [x] All commits have been signed for [DCO](https://developercertificate.org/) --------- Signed-off-by: Nahuel Espinosa <nespinosa@ekumenlabs.com> Co-authored-by: Nahuel Espinosa <nespinosa@ekumenlabs.com>
|
Addressed as of #275. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Feature description
Connected to #172. Good ol' AMCL computes sampled mean and covariance from particle clusters, whereas Beluga AMCL computes sampled mean and covariance for the entire posterior particle cloud. Regardless of any arguments in favor or against either approach, Beluga AMCL should offer the choice. We need another estimation mixin.
The text was updated successfully, but these errors were encountered: