update

Author: ShuoXing98

README.md

@@ -1,2 +1,15 @@
-# Re-Align
-A novel alignment framework that leverages image retrieval to mitigate hallucinations in Vision Language Models.
+<p align="center" width="60%">
+<img src="assets/pics/logo.png" alt="Re-Align" style="width: 35%; min-width: 200px; display: block; margin: auto; background-color: transparent;">
+</p>
+
+<div id="top" align="center">
+
+[![](https://img.shields.io/badge/Project%20Page-8A2BE2)](https://taco-group.github.io/AutoTrust/)
+![Code License](https://img.shields.io/badge/Code%20License-Apache%202.0-brightgreen)
+[![arXiv](https://img.shields.io/badge/arXiv-2412.15206-b31b1b.svg)](https://arxiv.org/abs/2412.15206)
+
+</div>
+
+# Re-Align: Aligning Vision Language Models via Retrieval-Augmented Direct Preference Optimization
+
+**Re-Align** is a novel alignment framework that leverages image retrieval to mitigate hallucinations in Vision Language Models.