Update README.md

README.md

@@ -1,10 +1,58 @@
 ---
-license: mit
-pipeline_tag: zero-shot-classification
+---
+license: apache-2.0
 tags:
-- model_hub_mixin
-- pytorch_model_hub_mixin
+- medical-imaging
+- vision-language-model
+- vlm
+- lora
+- graph-neural-networks
+- zero-shot
+datasets:
+- physionet/mimic-cxr-jpg
+metrics:
+- accuracy
 ---
 
+# ACE-LoRA: Graph-Attentive Context Enhancement for Medical VLMs
+
+<div align="center">
+  <a href="https://arxiv.org/pdf/2603.17079">
+    <img src="https://img.shields.io/badge/arXiv-2603.17079-b31b1b.svg" alt="arXiv">
+  </a>
+</div>
+
+**ACE-LoRA** is a parameter-efficient adaptation framework designed for generalist medical Vision-Language Models (VLMs). It addresses the specialization–generalization trade-off by integrating Low-Rank Adaptation (LoRA) with a novel **Attention-based Context Enhancement Hypergraph Neural Network (ACE-HGNN)**.
+
+## Model Description
+
+Existing medical VLMs often struggle to balance broad semantic understanding with fine-grained diagnostic cues. ACE-LoRA bridges this gap by adding only **$0.95M$** trainable parameters to frozen image-text encoders.
+
+### Key Features:
+* **ACE-HGNN Module:** Captures higher-order contextual interactions beyond pairwise similarity, enriching global representations with localized diagnostic details.
+* **Label-Guided InfoNCE Loss:** A specialized loss formulation designed to suppress false negatives between semantically related image-text pairs, improving cross-modal alignment.
+* **Efficiency:** Achieves state-of-the-art (SOTA) performance across multiple domains while keeping the backbone frozen.
+
+> [!NOTE]
+> **Abstract:** ACE-LoRA integrates LoRA modules into frozen image-text encoders and introduces a Hypergraph Neural Network to capture contextual interactions. Despite its minimal parameter footprint, it consistently outperforms SOTA medical VLMs and PEFT baselines in zero-shot classification, segmentation, and detection tasks.
+
+---
+
+## Technical Specifications
+
+### Architecture Overview
+The model utilizes a frozen CLIP-like backbone enhanced with:
+1.  **LoRA Adapters:** Plotted within the transformer layers of the vision and text encoders.
+2.  **ACE-HGNN:** A hypergraph-based module that processes localized features to capture complex diagnostic patterns.
+
+
+### Environment Setup
+The framework was developed using `Python 3.10.18` and `PyTorch 2.1.0` with `CUDA 11.8`.
+
+```bash
+conda create -n ace_lora python=3.10.18
+conda install pytorch==2.1.0 torchvision==0.16.0 torchaudio==2.1.0 pytorch-cuda=11.8 -c pytorch -c nvidia
+pip install -r requirements.txt
+
 - Code: https://github.com/icon-lab/ACE-LoRA
 - Paper: https://arxiv.org/pdf/2603.17079