Litian2002/robotics-papers-vecdb

---
license: mit
tags:
- robotics
- papers
- embeddings
- lancedb
- semantic-search
size_categories:
- 10K<n<100K
---

# Robotics Papers Vector Database

Semantic search over **63,381 academic papers** from 30 conference-year combinations in robotics, CV, and ML.

## Contents

- **Embeddings**: BAAI/bge-m3 (1024 dimensions) via SiliconFlow
- **Conferences**: CoRL, CVPR, ECCV, ICCV, ICLR, ICML, ICRA, IROS, NeurIPS, RSS, WACV (2023-2026)
- **Fields**: title, abstract, author, conference, year, arxiv, github, citations, keywords
- **Format**: LanceDB (compacted, single fragment)

## Usage

### Remote query (no download needed)

```python
import lancedb

db = lancedb.connect("hf://datasets/Litian2002/robotics-papers-vecdb/lancedb")
table = db.open_table("papers")
print(f"Papers: {table.count_rows()}")

# Semantic search (requires embedding your query first)
# See the companion repo for the full search pipeline
```

### Local usage

```bash
git clone https://huggingface.co/datasets/Litian2002/robotics-papers-vecdb
# or use huggingface_hub to download
```

### With vec-db CLI

```bash
# Clone the tool repo
git clone <vec-db-repo-url>
cd vec-db
VECDB_LANCE_DIR=/path/to/downloaded/lancedb npx tsx src/cli.ts search "robot grasping"
```

## Schema

| Column | Type | Description |
|--------|------|-------------|
| vecId | string | `{conf}_{year}_{id}` compound key |
| title | string | Paper title |
| abstract | string | Paper abstract |
| author | string | Authors (semicolon-separated) |
| conference | string | Conference name |
| year | float | Publication year |
| arxiv | string | arXiv ID |
| github | string | GitHub repo URL |
| gsCitation | float | Citation count (from OpenAlex) |
| vector | float[1024] | bge-m3 embedding |