update README

README.md

@@ -1,9 +1,11 @@
 <div align="center">
 
-# InternDataEngine: A simulation-based data generation engine designed for robotic learning. 
+# InternDataEngine: A simulation-based data generation engine designed for robotic learning.
 
 </div>
 
+<div align="center">
+
 [![Paper InternData-A1](https://img.shields.io/badge/Paper-InternData--A1-red.svg)](https://arxiv.org/abs/2511.16651)
 [![Paper Nimbus](https://img.shields.io/badge/Paper-Nimbus-red.svg)](https://arxiv.org/abs/2601.21449)
 [![Paper InternVLA-M1](https://img.shields.io/badge/Paper-InternVLA--M1-red.svg)](https://arxiv.org/abs/2510.13778)
@@ -11,16 +13,21 @@
 [![Data InternData-M1](https://img.shields.io/badge/Data-InternData--M1-blue?logo=huggingface)](https://huggingface.co/datasets/InternRobotics/InternData-M1)
 [![Docs](https://img.shields.io/badge/Docs-TBD-lightgrey.svg)](#)
 
+</div>
+
 ## 💻 About
 
-InternDataEngine is a data-centric engine for embodied AI that powers large-scale model training and iteration.  
-Built on NVIDIA Isaac Sim, it unifies high-fidelity physical interaction from InternData-A1, semantic task and scene generation from InternData-M1, and high-throughput scheduling from the Nimbus framework to deliver realistic, task-aligned, and massively scalable robotic manipulation data.
+<div align="center">
+  <img src="./docs/images/intern_data_engine.jpeg" alt="InternDataEngine Overview" width="80%">
+</div>
+
+InternDataEngine is a synthetic data generation engine for embodied AI that powers large-scale model training and iteration. Built on NVIDIA Isaac Sim, it unifies high-fidelity physical interaction from InternData-A1, semantic task and scene generation from InternData-M1, and high-throughput scheduling from the Nimbus framework to deliver realistic, task-aligned, and massively scalable robotic manipulation data.
 
 - **More realistic physical interaction**: Unified simulation of rigid, articulated, deformable, and fluid objects across single-arm, dual-arm, and humanoid robots, enabling long-horizon, skill-composed manipulation that better supports sim-to-real transfer.
 - **More task-aligned data generation**: LLM-driven task and instruction generation with task-oriented scene graphs (ToSG), producing structured scenes and rich multi-modal annotations (boxes, keypoints, trajectories) for complex instruction-following and spatial reasoning.
 - **More efficient large-scale production**: Nimbus-powered asynchronous pipelines that decouple planning, rendering, and storage, achieving 2–3× end-to-end throughput, cluster-level load balancing and fault tolerance for billion-scale data generation.
 
-## 🔥 Latest News 
+## 🔥 Latest News
 
 - **[2026/03]** We release the InternDataEngine codebase, which includes the core modules: InternData-A1, Nimbus, and InternData-M1.
 
@@ -55,3 +62,13 @@ All the code within this repo are under [CC BY-NC-SA 4.0](https://creativecommon
   year={2025}
 }
 ```
+
+<!--
+```BibTeX
+@misc{interndataengine2026,
+  title={InternDataEngine: A Synthetic Data Generation Engine for Robotic Learning},
+  author={InternDataEngine Contributors},
+  year={2026},
+  }
+}
+``` -->