Update README.md
Browse files
README.md
CHANGED
|
@@ -19,6 +19,8 @@ tags:
|
|
| 19 |
- Stable Diffusion v1-5
|
| 20 |
---
|
| 21 |
|
|
|
|
|
|
|
| 22 |
# OpenSDI-SD1.5-SigLIP2
|
| 23 |
|
| 24 |
> OpenSDI-SD1.5-SigLIP2 is a vision-language encoder model fine-tuned from google/siglip2-base-patch16-224 for binary image classification. It is trained to detect whether an image is a real photograph or generated using Stable Diffusion 1.5 (SD1.5), utilizing the SiglipForImageClassification architecture.
|
|
@@ -120,4 +122,4 @@ OpenSDI-SD1.5-SigLIP2 is designed for the following use cases:
|
|
| 120 |
* Generative Model Evaluation – Detect SD1.5-generated images for analysis and benchmarking.
|
| 121 |
* Dataset Integrity – Filter out AI-generated images from real-world image datasets.
|
| 122 |
* Digital Media Forensics – Support visual content verification and source validation.
|
| 123 |
-
* Trust & Safety – Detect synthetic media used in deceptive or misleading contexts.
|
|
|
|
| 19 |
- Stable Diffusion v1-5
|
| 20 |
---
|
| 21 |
|
| 22 |
+
|
| 23 |
+
|
| 24 |
# OpenSDI-SD1.5-SigLIP2
|
| 25 |
|
| 26 |
> OpenSDI-SD1.5-SigLIP2 is a vision-language encoder model fine-tuned from google/siglip2-base-patch16-224 for binary image classification. It is trained to detect whether an image is a real photograph or generated using Stable Diffusion 1.5 (SD1.5), utilizing the SiglipForImageClassification architecture.
|
|
|
|
| 122 |
* Generative Model Evaluation – Detect SD1.5-generated images for analysis and benchmarking.
|
| 123 |
* Dataset Integrity – Filter out AI-generated images from real-world image datasets.
|
| 124 |
* Digital Media Forensics – Support visual content verification and source validation.
|
| 125 |
+
* Trust & Safety – Detect synthetic media used in deceptive or misleading contexts.
|