---
pretty_name: 3DModelSearch Renders & Embeddings
task_categories:
  - feature-extraction
tags:
  - 3d
  - retrieval
  - multimodal
  - objaverse
  - languagebind
size_categories:
  - 100B<n<1T
license: other
license_name: mixed-see-provenance
---

# 3DModelSearch — Renders & Embeddings

Precomputed data for [3DModelSearch](<https://github.com/srutisrinidhi/3DModelSearch>): multi-view image renders,
turntable videos, and LanguageBind embeddings for 3D models, used in Learning Visual and Motion Aware 3D Model Representations
for Semantic Retriev.
These let you reproduce retrieval results without re-rendering or re-embedding.

## Contents

| File | Size | Contents |
| --- | --- | --- |
| `embeddings.tar.gz` | ~<X> GB | `out/Embeddings/<uid>_image.pt`, `<uid>_video.pt`, and `out/models_data.csv` |
| `videos.tar.gz` | ~17 GB | `data/Videos/<shard>/<uid>.mp4` |
| `images.tar.gz` | ~197 GB | `data/Images/<shard>/<uid>/view0..5.jpg` |

- **Image embeddings**: `[6, 768]` tensors (6 views, LanguageBind image encoder).
- **Video embeddings**: `[768]` tensors (LanguageBind video encoder).
- `models_data.csv` maps each `uid` to its render + embedding paths (**paths are relative to
  the code repo root**).

Source `.glb` models are **not** included — fetch them from
[Objaverse](https://objaverse.allenai.org/) as described in the code repo.

## Usage

```bash
pip install huggingface_hub
huggingface-cli download <HF_USER>/<REPO> embeddings.tar.gz --repo-type dataset --local-dir .
tar -xzf embeddings.tar.gz     # -> out/
tar -xzf videos.tar.gz         # -> data/Videos/
tar -xzf images.tar.gz         # -> data/Images/