File size: 3,367 Bytes
d536e7b
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
ef7c7fc
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
d536e7b
ef7c7fc
 
 
 
 
756d400
 
 
 
 
 
 
 
 
 
 
c57c129
 
435c5b2
c6ed7ea
435c5b2
66aeb71
756d400
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
---
dataset_info:
  features:
  - name: prompt
    dtype: string
  - name: image1
    dtype: image
  - name: image2
    dtype: image
  - name: model1
    dtype: string
  - name: model2
    dtype: string
  - name: weighted_score_image1
    dtype: float32
  - name: weighted_score_image2
    dtype: float32
  - name: detailed_results
    dtype: string
  splits:
  - name: train
    num_bytes: 1691529922.55
    num_examples: 2310
  download_size: 1513490728
  dataset_size: 1691529922.55
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
license: cdla-permissive-2.0
task_categories:
- text-to-image
- image-to-text
- image-classification
- reinforcement-learning
language:
- en
tags:
- Human
- Preference
- Coherence
- Alignment
- country
- language
- bytedance
- seedream-3
- qwen
- qwen-image
- recraft-ai
- recraft-v3
- minimax
- image-01
- google
- imagen-4-fast
- black-forest-labs
- flux
- prunaai
- wan-2.2-image
- ideogram-ai
- ideogram
- luma
- photon
- bria
- image-3.2
- luma
- photon-flash
- black-forest-labs
- flux
size_categories:
- 10K<n<100K
pretty_name: Human Face Generation Alignment
---

# Rapidata Human Face Generation Alignment

<a href="https://www.rapidata.ai">
<img src="https://cdn-uploads.huggingface.co/production/uploads/66f5624c42b853e73e0738eb/jfxR79bOztqaC6_yNNnGU.jpeg" width="400" alt="Dataset visualization">
</a>

This T2I dataset contains over ~22'000 human responses, collected in less than 1h using the [Rapidata Python API](https://docs.rapidata.ai), accessible to anyone and ideal for large scale evaluation.
Evaluating 12 different image generation models on which one can generate faces more accurately.

The question that the annotators get asked is: "Which Image follows the description of the human better?"

<strong>To evaluate your own models and create leaderboard check out our [MRI](https://docs.rapidata.ai/mri/) </strong>

## The resulting leaderboard

Checkout the full leaderboard here: https://app.rapidata.ai/mri/benchmarks/68af24ae74482280b62f7596

<a href="https://app.rapidata.ai/mri/benchmarks/68af24ae74482280b62f7596">
<img src="https://cdn-uploads.huggingface.co/production/uploads/66f5624c42b853e73e0738eb/ABLvkbrkTYBNRzipiFJNs.png" alt="Leaderboard">
</a>

|    | name                              | wins | total_matches | score      |
|----|-----------------------------------|------|---------------|------------|
| 0  | bytedance/seedream-3              | 2604 | 4472          | 1123.630000|
| 1  | qwen/qwen-image                   | 1654 | 2877          | 1110.540000|
| 2  | recraft-ai/recraft-v3             | 1157 | 2028          | 1105.070000|
| 3  | minimax/image-01                  | 2726 | 5125          | 1051.620000|
| 4  | google/imagen-4-fast              | 3205 | 6180          | 1027.690000|
| 5  | black-forest-labs/flux-1.1-pro    | 1272 | 2650          | 969.560000 |
| 6  | prunaai/wan-2.2-image             | 1121 | 2344          | 967.050000 |
| 7  | ideogram-ai/ideogram-v3-balanced  | 2868 | 6170          | 962.890000 |
| 8  | luma/photon                       | 1839 | 3934          | 955.450000 |
| 9  | bria/image-3.2                    | 1507 | 3226          | 951.530000 |
| 10 | luma/photon-flash                 | 1499 | 3529          | 889.910000 |
| 11 | black-forest-labs/flux-schnell    | 1024 | 2417          | 885.060000 |