Datasets:

Modalities:
Image
Text
Formats:
parquet
Languages:
English
ArXiv:
Libraries:
Datasets
Dask
License:
hellcaster commited on
Commit
64cc56f
·
verified ·
1 Parent(s): a49fe61

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +94 -3
README.md CHANGED
@@ -1,4 +1,5 @@
1
  ---
 
2
  dataset_info:
3
  features:
4
  - name: image
@@ -7,13 +8,103 @@ dataset_info:
7
  dtype: string
8
  splits:
9
  - name: train
10
- num_bytes: 653490752.49
11
  num_examples: 1127
12
- download_size: 618158413
13
- dataset_size: 653490752.49
14
  configs:
15
  - config_name: default
16
  data_files:
17
  - split: train
18
  path: data/train-*
 
 
 
 
 
 
 
 
 
 
 
19
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
+ license: apache-2.0
3
  dataset_info:
4
  features:
5
  - name: image
 
8
  dtype: string
9
  splits:
10
  - name: train
11
+ num_bytes: 653490753.125
12
  num_examples: 1127
13
+ download_size: 618157948
14
+ dataset_size: 653490753.125
15
  configs:
16
  - config_name: default
17
  data_files:
18
  - split: train
19
  path: data/train-*
20
+ task_categories:
21
+ - object-detection
22
+ language:
23
+ - en
24
+ tags:
25
+ - accessibility
26
+ - macOS
27
+ - hierarchy
28
+ pretty_name: Screen2AX-Tree
29
+ size_categories:
30
+ - 1K<n<10K
31
  ---
32
+ # 📦 Screen2AX-Tree
33
+
34
+ Screen2AX-Tree is part of the **Screen2AX** dataset suite, a research-driven collection for advancing accessibility in macOS applications using computer vision and deep learning.
35
+
36
+ This dataset provides **hierarchical accessibility annotations** of macOS application screenshots, structured as serialized trees. It is designed for training models that reconstruct accessibility hierarchies from visual input.
37
+
38
+ ---
39
+
40
+ ## 🧠 Dataset Summary
41
+
42
+ Each sample in the dataset consists of:
43
+ - An application **screenshot** (`image`)
44
+ - A serialized **accessibility tree** (`accessibility`): A JSON-formatted string representing the UI structure, including roles, bounds, and child relationships.
45
+
46
+ **Task Category:**
47
+ - `object-detection` (structured / hierarchical)
48
+
49
+ **Language:**
50
+ - English (`en`)
51
+
52
+ ---
53
+
54
+ ## 📚 Usage
55
+
56
+ ### Load with `datasets` library
57
+
58
+ ```python
59
+ from datasets import load_dataset
60
+
61
+ dataset = load_dataset("MacPaw/Screen2AX-Tree")
62
+ ```
63
+
64
+ ### Example structure
65
+
66
+ ```python
67
+ sample = dataset["train"][0]
68
+ print(sample.keys())
69
+ # dict_keys(['image', 'accessibility'])
70
+
71
+ print(sample["accessibility"])
72
+ # '{ "role": "AXWindow", "children": [ ... ] }'
73
+ ```
74
+
75
+ You can parse the accessibility field as JSON to work with the structured hierarchy:
76
+
77
+ ```python
78
+ import json
79
+
80
+ tree = json.loads(sample["accessibility"])
81
+ ```
82
+
83
+ ---
84
+
85
+ ## 📜 License
86
+
87
+ This dataset is licensed under the **Apache 2.0 License**.
88
+
89
+ ---
90
+
91
+ ## 🔗 Related Projects
92
+
93
+ - [Screen2AX Main Project Page](https://github.com/MacPaw/Screen2AX)
94
+ - [Screen2AX HuggingFace Collection](https://huggingface.co/collections/MacPaw/screen2ax-687dfe564d50f163020378b8)
95
+
96
+ ---
97
+
98
+ ## ✍️ Citation
99
+
100
+ If you use this dataset, please cite the Screen2AX paper:
101
+
102
+ ```bibtex
103
+ ...
104
+ ```
105
+
106
+ ---
107
+
108
+ ## 🌐 MacPaw Research
109
+
110
+ Learn more at [https://research.macpaw.com](https://research.macpaw.com)