narugo commited on
Commit
32d6c8b
1 Parent(s): fbcf2a3

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +40 -0
README.md CHANGED
@@ -7,10 +7,50 @@ tags:
7
  - art
8
  datasets:
9
  - deepghs/ai_image_corrupted
 
10
  ---
11
 
 
 
 
 
 
 
12
  | Name | FLOPS | Params | Accuracy | AUC | Confusion | Labels |
13
  |:-------------------------:|:-------:|:--------:|:----------:|:------:|:---------------------------------------------------------------------------------------------------------------------:|:---------------------:|
14
  | caformer_s36_v0_focal | 22.10G | 37.21M | 95.65% | 0.9916 | [confusion](https://huggingface.co/deepghs/ai_image_corrupted/blob/main/caformer_s36_v0_focal/plot_confusion.png) | `corrupted`, `normal` |
15
  | caformer_s36_v0_sce | 22.10G | 37.21M | 95.77% | 0.9894 | [confusion](https://huggingface.co/deepghs/ai_image_corrupted/blob/main/caformer_s36_v0_sce/plot_confusion.png) | `corrupted`, `normal` |
16
  | mobilenetv3_v0_focal_dist | 0.63G | 4.18M | 94.02% | 0.9842 | [confusion](https://huggingface.co/deepghs/ai_image_corrupted/blob/main/mobilenetv3_v0_focal_dist/plot_confusion.png) | `corrupted`, `normal` |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
7
  - art
8
  datasets:
9
  - deepghs/ai_image_corrupted
10
+ library_name: dghs-imgutils
11
  ---
12
 
13
+ This is the classifier model for predicting the anime-style stable-diffusion-generated images are corrupted or not.
14
+
15
+ Trained on dataset [deepghs/ai_image_corrupted](https://huggingface.co/datasets/deepghs/ai_image_corrupted).
16
+
17
+ ## Models
18
+
19
  | Name | FLOPS | Params | Accuracy | AUC | Confusion | Labels |
20
  |:-------------------------:|:-------:|:--------:|:----------:|:------:|:---------------------------------------------------------------------------------------------------------------------:|:---------------------:|
21
  | caformer_s36_v0_focal | 22.10G | 37.21M | 95.65% | 0.9916 | [confusion](https://huggingface.co/deepghs/ai_image_corrupted/blob/main/caformer_s36_v0_focal/plot_confusion.png) | `corrupted`, `normal` |
22
  | caformer_s36_v0_sce | 22.10G | 37.21M | 95.77% | 0.9894 | [confusion](https://huggingface.co/deepghs/ai_image_corrupted/blob/main/caformer_s36_v0_sce/plot_confusion.png) | `corrupted`, `normal` |
23
  | mobilenetv3_v0_focal_dist | 0.63G | 4.18M | 94.02% | 0.9842 | [confusion](https://huggingface.co/deepghs/ai_image_corrupted/blob/main/mobilenetv3_v0_focal_dist/plot_confusion.png) | `corrupted`, `normal` |
24
+
25
+ ## How to Use
26
+
27
+ You can use this model with [dghs-imgutils](https://dghs-imgutils.deepghs.org/main/api_doc/generic/classify.html).
28
+
29
+ ```python
30
+ from imgutils.generic import classify_predict_score
31
+
32
+ classify_predict_score(
33
+ 'sample_image.png',
34
+ repo_id='deepghs/ai_image_corrupted',
35
+ model_name='mobilenetv3_v0_focal_dist',
36
+ )
37
+ # {'corrupted': 0.7807788848876953, 'normal': 0.2192210853099823}
38
+ ```
39
+
40
+ ## Note
41
+
42
+ This model is trained on SD1.5 images, generated by 8 different base models.
43
+
44
+ So the better solution for this problem is to use some metrics like artstyle embeddings,
45
+ this model will be deprecated as soon as the artstyle embedding model for anime images is completed.
46
+
47
+ ## Citation
48
+
49
+ ```
50
+ @misc{Citation,
51
+ title={AI-Corrupt Score for Anime Images},
52
+ author={narugo1992},
53
+ year={2023},
54
+ howpublished={\url{https://huggingface.co/deepghs/ai_image_corrupted}}
55
+ }
56
+ ```