EFM3D Scene Embeddings (background)

1 EFM3D Scene Embeddings

This page retains cited background on EFM3D/EVL and Project Aria assets [1]. It is not a schema or implementation owner. The thesis defines the actor-visible information boundary; Python docstrings, config, and tests define readers, caches, and validation. Planned representations must not be inferred from this navigation page.

See EFM3D/EVL literature, ASE dataset, and semi-dense point clouds. The canonical project discussion is in the Typst thesis.

References

[1]

J. Straub, D. DeTone, T. Shen, N. Yang, C. Sweeney, and R. Newcombe, “EFM3D: A benchmark for measuring progress towards 3D egocentric foundation models.” 2024. Available: https://arxiv.org/abs/2406.10224

---
title: "EFM3D Scene Embeddings (background)"
audience: public
phase: archive
status: deprecated
owner: docs
format: html
bibliography: ../../references.bib
---

# EFM3D Scene Embeddings {#efm3d-scene-embeddings}

This page retains cited background on EFM3D/EVL and Project Aria assets
[@EFM3D-straub2024]. It is not a schema or implementation owner. The thesis
defines the actor-visible information boundary; Python docstrings, config, and
tests define readers, caches, and validation. Planned representations must not
be inferred from this navigation page.

See [EFM3D/EVL literature](../literature/efm3d.qmd), [ASE dataset](../ase_dataset.qmd),
and [semi-dense point clouds](semi-dense-pc.qmd). The canonical project
discussion is in the [Typst thesis](../../typst/thesis/main.typ).