File size: 3,873 Bytes
7e1698a
4f775e4
 
 
 
 
 
 
 
 
 
 
 
 
7e1698a
4f775e4
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
e621988
 
4f775e4
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
f7c30e7
4f775e4
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
e621988
 
 
 
4f775e4
 
 
 
 
f7c30e7
4f775e4
 
 
 
 
 
 
 
 
 
 
 
 
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
---
language:
- en
size_categories:
- 100K<n<1M
task_categories:
- image-classification
pretty_name: DPChallenge Photo Metadata Dataset
tags:
- photography
- image-quality
- human-labels
- ratings
- Evaluation
---

# DPChallenge Photo Metadata Dataset

## Dataset Description

This dataset contains metadata and statistics from DPChallenge, a photography community platform where photographers participate in themed challenges and receive peer ratings. 

**Key Features:**
- **Valuable Human Labels**: Contains human-scored quality ratings from multiple rater groups (all users, commenters, participants, non-participants)
- **Collection Date**: Dec 2025
- **Data Quality**: 
  - Only includes images with complete statistics
  - Scores validated and corrected for parsing errors (original scrape had ~1,400 rows with incorrectly parsed scores from user comments)
  - Score range: 2.0 - 8.6 (realistic competition scores)

**Note**: In all columns, a value of `0` represents N/A (not available/not applicable).

## Dataset Structure

### Data Splits

- **Train**: 184,895 images (90%)
- **Test**: 20,544 images (10%)

The test set is stratified by score bins (0-4, 5-6, 7-10) to ensure even distribution of image quality ratings.

### Features

- `image_id` (int): Unique identifier for the image
- `gallery_id` (int): Challenge/gallery identifier
- `gallery_name` (string): Name of the challenge/gallery
- `title` (string): Photo title
- `photographer` (string): Photographer username
- `challenge` (string): Challenge theme
- `editing_rule` (string): Editing rules (Basic/Advanced)
- `camera` (string): Camera and lens information
- `location` (string): Photo location
- `date_uploaded` (string): Upload date
- `place` (string): Additional location details
- `total_entries` (int): Total entries in the challenge
- `avg_all_users` (float): Average rating from all users (0-10 scale)
- `avg_commenters` (float): Average rating from commenters
- `avg_participants` (float): Average rating from challenge participants
- `avg_nonparticipants` (float): Average rating from non-participants
- `views_since_voting` (int): Views after voting period
- `views_during_voting` (int): Views during voting period
- `votes` (int): Number of votes received
- `comments_count` (int): Number of comments
- `favorites_count` (int): Number of times favorited
- `image` (Image): The actual image file
- `has_statistics` (bool): Whether full statistics are available

### Score Distribution

The dataset uses stratified sampling based on rounded `avg_all_users` scores (the main quality metric):
- **0-4**: Low-rated images
- **5-6**: Medium-rated images  
- **7-8+**: High-rated images (max score in dataset is ~8.6)

Most scores fall in the 4-7 range, with very few extreme scores. This reflects real-world photography competition dynamics where truly exceptional images are rare.

### Rating Statistics

Statistics for human label scores across the dataset:

| Metric | Min | Max | Mean | Std |
|--------|-----|-----|------|-----|
| `avg_all_users` | 2.00 | 8.60 | 5.45 | 0.73 |
| `avg_commenters` | 1.00 | 10.00 | 6.28 | 1.32 |
| `avg_participants` | 1.77 | 9.33 | 5.29 | 0.75 |
| `avg_nonparticipants` | 1.96 | 8.62 | 5.57 | 0.74 |

**Note**: `avg_all_users` is used as the primary splitting criterion for stratified train/test split to ensure balanced quality distribution.

## Source Data

Data scraped from [DPChallenge](https://www.dpchallenge.com/) public photo galleries. This dataset includes both images and metadata.


## Citation

If you use this dataset, please cite DPChallenge as the source of the data:
```
DPChallenge Photography Community
https://www.dpchallenge.com/
```

## License

This dataset contains metadata from publicly available information on DPChallenge. Images remain the property of their respective photographers. Please respect copyright and usage terms.