Update README.md
Browse files
README.md
CHANGED
|
@@ -29,12 +29,13 @@ license: mit
|
|
| 29 |
</p>
|
| 30 |
|
| 31 |
## 📖 Introduction
|
| 32 |
-
|
| 33 |
**UniPic3-Consistency-Model** is a few-step image editing and multi-image composition model based on **Consistency Flow Matching (CM)**.
|
| 34 |
The model learns a *trajectory-consistent* mapping from noisy latent states to clean images, enabling stable generation with strong structural consistency.
|
| 35 |
It is distilled from **UniPic-3** to support **fast inference (≤8 steps)** while preserving composition correctness.The model is especially suitable for scenarios requiring **geometric alignment** and **semantic coherence**, such as multi-image composition and human–object interaction (HOI).
|
| 36 |
|
| 37 |
## 📊 Benchmarks
|
|
|
|
| 38 |
|
| 39 |
|
| 40 |
## 🧠 Usage
|
|
|
|
| 29 |
</p>
|
| 30 |
|
| 31 |
## 📖 Introduction
|
| 32 |
+
<div align="center"> <img src="unipic3.png" alt="Model Teaser" width="720"> </div>
|
| 33 |
**UniPic3-Consistency-Model** is a few-step image editing and multi-image composition model based on **Consistency Flow Matching (CM)**.
|
| 34 |
The model learns a *trajectory-consistent* mapping from noisy latent states to clean images, enabling stable generation with strong structural consistency.
|
| 35 |
It is distilled from **UniPic-3** to support **fast inference (≤8 steps)** while preserving composition correctness.The model is especially suitable for scenarios requiring **geometric alignment** and **semantic coherence**, such as multi-image composition and human–object interaction (HOI).
|
| 36 |
|
| 37 |
## 📊 Benchmarks
|
| 38 |
+
<div align="center"> <img src="unipic3_eval.png" alt="Model Teaser" width="720"> </div>
|
| 39 |
|
| 40 |
|
| 41 |
## 🧠 Usage
|