Add model card
#1
by nielsr HF Staff - opened
README.md
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
pipeline_tag: image-to-image
|
| 3 |
+
---
|
| 4 |
+
|
| 5 |
+
# TextRIRO
|
| 6 |
+
|
| 7 |
+
TextRIRO is a conditional diffusion model for scene text editing: it changes textual content while preserving the original style. It is presented in [Real In, Real Out: What If We Only Use Real Data for Scene Text Editing?](https://huggingface.co/papers/2610.27592).
|
| 8 |
+
|
| 9 |
+
The approach uses a "Shuffle&Crop" strategy to construct self-paired training data from real images, avoiding synthetic paired STE training data. Code and usage instructions are available in the [GitHub repository](https://github.com/YesianRohn/TextRIRO).
|