An opinionated atlas of 6D object pose research

What it does

This is a curated index of research on 6D object pose estimation—locating an object in space and determining its orientation—plus related work on single-view 3D reconstruction and hand-object pose estimation. The maintainer sorts papers by venue and year, spanning CVPR 2014 through CVPR 2024, and includes arXiv preprints, journal articles, datasets, workshops, and a directory of researchers by region. A handful of open-source demos like CenterSnap and BundleTrack are linked, but the repository itself is strictly a reading list.

The interesting bit

The scope is intentionally narrow: geometry-based SFM and SLAM work is explicitly excluded because the maintainer has no interest in it, which gives the list a deep-learning-and-vision slant rather than claiming to be exhaustive. That editorial stance, combined with regional researcher directories and dedicated BOP Challenge sections, makes it feel more like a community field guide than an algorithm catalog.

Key highlights

• Covers roughly 2014–2024 across CVPR, ICCV, ECCV, ICRA, IROS, TPAMI, and IJCV.
• Includes adjacent topics like single-view 3D reconstruction and 3D hand-object pose estimation.
• Maintains sections for datasets, workshops, and BOP Challenge benchmarks.
• Flags recent additions with a 🔥 emoji so new entries stand out.
• Explicitly excludes SFM/SLAM papers, redirecting those to a separate awesome list.

Caveats

• The README is a single long scroll with minimal categorization beyond year and venue; finding a specific method still requires manual scanning.
• This is a bibliography, not a framework—there is no code, comparison tables, or unified evaluation script inside.

Verdict

Bookmark it if you are doing a literature review in object pose estimation or need a quick pulse on what the major vision and robotics venues published recently. Skip it if you want runnable code, unified benchmarks, or SLAM-based geometry work.