---
license: mit
language:
- en
tags:
- 3d-scene-understanding
- cross-modal-alignment
- CVPR-2025
- cvpr
---

This repo contains the pre-trained checkpoints released for: "CrossOver: 3D Scene Cross-Modal Alignment" (CVPR 2025, Highlight)

[Paper](https://arxiv.org/abs/2502.15011) | [Project Page](https://sayands.github.io/crossover/) | [Code](https://github.com/GradientSpaces/CrossOver/)

<p align="center">
  <a href="">
    <img src="https://sayands.github.io/crossover/static/videos/teaser.gif" width="100%">
  </a>
</p>