| Submission name | |||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Submission time (UTC) | Sept. 18, 2024, 9:29 a.m. | ||||||||||||||
| User | sp9103 | ||||||||||||||
| Task | Model-based 6D detection of unseen objects | ||||||||||||||
| Dataset | LM-O | ||||||||||||||
| Description | |||||||||||||||
| Evaluation scores |
|
| User | sp9103 |
|---|---|
| Publication | Nguyen et al.: GigaPose: Fast and Robust Novel Object Pose Estimation via One Correspondence, CVPR 2024. |
| Implementation | https://github.com/nv-nguyen/gigapose |
| Training image modalities | RGB-D |
| Test image modalities | RGB |
| Description | GigaPose (Nguyen et al., CVPR 2024) is a "hybrid" template-patch correspondence approach to estimate 6D pose of novel objects in RGB images: GigaPose first uses 162 templates, rendered images of the CAD models, to recover the out-of-plane rotation (2DoF) and then uses patch correspondences to estimate the remaining 4DoF. Detection method : CNOS-FastSAM |
| Computer specifications | RTX 3090Ti |