Update README.md

README.md

@@ -11,18 +11,19 @@ tags:
 - fiftyone
 - image
 - object-detection
-dataset_summary: '
+dataset_summary: >
 
 
 
 
-  This is a [FiftyOne](https://github.com/voxel51/fiftyone) dataset with 739 samples.
+  This is a [FiftyOne](https://github.com/voxel51/fiftyone) dataset with 739
+  samples.
 
 
   ## Installation
 
 
-  If you haven''t already, install FiftyOne:
+  If you haven't already, install FiftyOne:
 
 
   ```bash
@@ -44,9 +45,9 @@ dataset_summary: '
 
   # Load the dataset
 
-  # Note: other available arguments include ''max_samples'', etc
+  # Note: other available arguments include 'max_samples', etc
 
-  dataset = load_from_hub("harpreetsahota/FiftyOne-GUI-Grounding-Train")
+  dataset = load_from_hub("Voxel51/FiftyOne-GUI-Grounding-Train")
 
 
   # Launch the App
@@ -54,171 +55,196 @@ dataset_summary: '
   session = fo.launch_app(dataset)
 
   ```
-
-  '
+license: apache-2.0
 ---
 
-# Dataset Card for fiftyone_gui_train_set
-
-<!-- Provide a quick summary of the dataset. -->
-
-
-
-
-
-This is a [FiftyOne](https://github.com/voxel51/fiftyone) dataset with 739 samples.
-
-## Installation
-
-If you haven't already, install FiftyOne:
-
-```bash
-pip install -U fiftyone
-```
-
-## Usage
-
-```python
-import fiftyone as fo
-from fiftyone.utils.huggingface import load_from_hub
-
-# Load the dataset
-# Note: other available arguments include 'max_samples', etc
-dataset = load_from_hub("harpreetsahota/FiftyOne-GUI-Grounding-Train")
-
-# Launch the App
-session = fo.launch_app(dataset)
-```
 
+# Dataset Card for FiftyOne GUI Grounding Training Set
 
 ## Dataset Details
 
 ### Dataset Description
 
-<!-- Provide a longer summary of what this dataset is. -->
-
-
+This dataset contains 739 annotated GUI screenshots designed for training computer vision models to understand and interact with graphical user interfaces. The dataset uses the specialized COCO4GUI format, which extends the standard COCO detection format to handle GUI-specific features, interaction sequences, and rich metadata.
 
-- **Curated by:** [More Information Needed]
-- **Funded by [optional]:** [More Information Needed]
-- **Shared by [optional]:** [More Information Needed]
-- **Language(s) (NLP):** en
-- **License:** [More Information Needed]
+The dataset captures real user interface interactions across various applications and platforms, with both bounding box annotations for UI elements and precise keypoint annotations for interaction locations (clicks, taps, etc.).
 
-### Dataset Sources [optional]
+- **Curated by:** GUI Dataset Collector Tool
+- **Funded by:** [More Information Needed]
+- **Shared by:** Harpreet Sahota
+- **Language(s):** English (en)
+- **License:** Apache 2.0
 
-<!-- Provide the basic links for the dataset. -->
+### Dataset Sources
 
-- **Repository:** [More Information Needed]
-- **Paper [optional]:** [More Information Needed]
-- **Demo [optional]:** [More Information Needed]
+- **Repository:** [GUI Annotation Tool](https://github.com/harpreetsahota204/gui_annotation_tool)
+- **Paper:** [More Information Needed]
+- **Demo:** [More Information Needed]
 
 ## Uses
 
-<!-- Address questions around how the dataset is intended to be used. -->
-
 ### Direct Use
 
-<!-- This section describes suitable use cases for the dataset. -->
+This dataset is designed for:
 
-[More Information Needed]
+- **GUI Element Detection**: Training models to identify and locate UI components in screenshots
+- **Interaction Point Prediction**: Learning to predict where users should click/interact within interfaces  
+- **GUI Understanding**: Developing models that can comprehend user interface layouts and functionality
+- **Workflow Sequence Learning**: Understanding sequences of user interactions and task flows
+- **Cross-platform GUI Analysis**: Training on diverse applications and platforms (macOS, various browsers, etc.)
 
 ### Out-of-Scope Use
 
-<!-- This section addresses misuse, malicious use, and uses that the dataset will not work well for. -->
-
-[More Information Needed]
+- **Privacy-sensitive Applications**: Screenshots may contain application interfaces that could reveal user data
+- **Real-time Automation**: This is a training dataset, not intended for direct deployment in production automation systems
+- **Non-GUI Computer Vision**: Optimized specifically for GUI interactions, not general object detection
 
 ## Dataset Structure
 
-<!-- This section provides a description of the dataset fields, and additional information about the dataset structure such as criteria used to create the splits, relationships between data points, etc. -->
-
-[More Information Needed]
+### Data Format
+- **Images**: 739 PNG screenshots with timestamps (2025-07-22 to 2025-08-08)  
+- **Annotations**: COCO4GUI format JSON with extended metadata
+- **Resolution**: Variable (typical: 2992x1866 pixels for macOS screenshots)
+
+### Annotation Categories
+The dataset includes 7 interaction types:
+1. **click** - Standard mouse clicks
+2. **type** - Text input actions  
+3. **select** - Selection operations
+4. **hover** - Mouse hover interactions
+5. **drag** - Drag and drop operations
+6. **right_click** - Context menu interactions
+7. **double_click** - Double-click actions
+
+### Annotation Structure
+Each annotation contains:
+- **Bounding boxes**: Rectangular regions around UI elements
+- **Keypoints**: Precise interaction coordinates (x, y, visibility)
+- **Rich attributes**: 
+  - `task_description`: Human-readable description of the interaction
+  - `action_type`: Type of interaction being performed
+  - `element_info`: Information about the UI element
+  - `custom_metadata`: Additional context-specific data
+
+### Metadata Fields
+- **Application**: Source application (Arc Browser, Chrome, etc.)
+- **Platform**: Operating system (primarily macOS)
+- **Date captured**: Timestamp of screenshot capture
+- **Sequence information**: For tracking multi-step workflows
+- **GUI metadata**: Application-specific context
 
 ## Dataset Creation
 
 ### Curation Rationale
 
-<!-- Motivation for the creation of this dataset. -->
-
-[More Information Needed]
+This dataset was created to address the need for high-quality, annotated GUI interaction data for training computer vision models that can understand and interact with user interfaces. The focus on both spatial (bounding boxes) and precise (keypoints) annotations enables training of models that can both identify UI elements and predict exact interaction points.
 
 ### Source Data
 
-<!-- This section describes the source data (e.g. news text and headlines, social media posts, translated sentences, ...). -->
-
 #### Data Collection and Processing
 
-<!-- This section describes the data collection and processing process such as data selection criteria, filtering and normalization methods, tools and libraries used, etc. -->
+Data was collected using a specialized GUI annotation tool that:
+- **Captures live screenshots** of various applications and interfaces
+- **Provides real-time annotation capabilities** with bounding boxes and click points
+- **Maintains rich metadata** about applications, platforms, and interaction context
+- **Supports sequence tracking** for multi-step user workflows
+- **Exports in COCO4GUI format** optimized for FiftyOne integration
 
-[More Information Needed]
+Screenshots span from July 22, 2025 to August 8, 2025, capturing diverse GUI states and interactions primarily on macOS systems using various browsers and applications.
 
 #### Who are the source data producers?
 
-<!-- This section describes the people or systems who originally created the data. It should also include self-reported demographic or identity information for the source data creators if this information is available. -->
+Screenshots were captured from real application interfaces during normal usage, with manual annotation of interaction points and UI elements using the GUI Dataset Collector tool.
 
-[More Information Needed]
-
-### Annotations [optional]
-
-<!-- If the dataset contains annotations which are not part of the initial data collection, use this section to describe them. -->
+### Annotations
 
 #### Annotation process
 
-<!-- This section describes the annotation process such as annotation tools used in the process, the amount of data annotated, annotation guidelines provided to the annotators, interannotator statistics, annotation validation, etc. -->
+Annotations were created using a specialized web-based GUI annotation tool that provides:
+- **Interactive bounding box drawing** for UI element regions
+- **Precise click point placement** for interaction coordinates  
+- **Rich metadata entry** for task descriptions and element information
+- **Real-time preview** and editing capabilities
+- **Sequence tracking** for multi-step workflows
 
-[More Information Needed]
+The tool automatically captures metadata like application name, platform, and timestamps while allowing annotators to add detailed descriptions of interactions and UI elements.
 
 #### Who are the annotators?
 
-<!-- This section describes the people or systems who created the annotations. -->
+Annotations were created by the dataset curator using the GUI annotation tool during real interface interactions.
 
-[More Information Needed]
+### Personal and Sensitive Information
 
-#### Personal and Sensitive Information
+Screenshots may contain application interfaces that could potentially reveal:
+- **Application usage patterns**
+- **Interface layouts and designs** 
+- **Partial text content** from applications
 
-<!-- State whether the dataset contains data that might be considered personal, sensitive, or private (e.g., data that reveals addresses, uniquely identifiable names or aliases, racial or ethnic origins, sexual orientations, religious beliefs, political opinions, financial or health data, etc.). If efforts were made to anonymize the data, describe the anonymization process. -->
-
-[More Information Needed]
+Care should be taken when using this dataset to ensure compliance with privacy requirements. The dataset focuses on interaction patterns rather than content, but users should review samples for any sensitive information relevant to their use case.
 
 ## Bias, Risks, and Limitations
 
-<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+### Technical Limitations
+- **Platform bias**: Primarily macOS screenshots, limiting cross-platform generalization
+- **Application bias**: Heavy focus on browser interfaces and specific applications
+- **Resolution consistency**: Variable screenshot resolutions may affect model training
+- **Temporal scope**: Data collected over a limited time period (July-August 2025)
+
+### Potential Biases
+- **Interface design bias**: Reflects specific application UI paradigms and design patterns
+- **Interaction pattern bias**: May not represent all user interaction styles or accessibility needs
+- **Language bias**: Primarily English-language interfaces
 
-[More Information Needed]
+### Risks
+- **Privacy concerns**: Screenshots may inadvertently contain sensitive interface elements
+- **Overfitting risk**: Limited diversity in applications and platforms
+- **Generalization challenges**: Models trained on this data may not generalize well to significantly different GUI paradigms
 
 ### Recommendations
 
-<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+Users should:
+- **Augment with additional data** from diverse platforms and applications
+- **Validate on target use cases** before deployment
+- **Consider privacy implications** when using in production systems
+- **Test cross-platform performance** if deploying across different operating systems
+- **Implement proper data handling** procedures for any sensitive content
 
-Users should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.
+## Technical Details
 
-## Citation [optional]
+### FiftyOne Integration
+This dataset includes advanced FiftyOne features:
+- **Brain embeddings**: CLIP and image similarity indices computed
+- **Visualization**: UMAP embeddings for data exploration  
+- **Uniqueness scoring**: Identifies diverse vs. redundant samples
+- **Representativeness metrics**: Helps identify core vs. outlier samples
 
-<!-- If there is a paper or blog post introducing the dataset, the APA and Bibtex information for that should go in this section. -->
+### Enhanced Metadata
+The COCO4GUI format provides:
+- **Sequence tracking**: Links related interaction steps
+- **Rich attributes**: Detailed interaction context
+- **GUI-specific fields**: Application, platform, timing metadata
+- **Workflow information**: Multi-step task sequences
 
-**BibTeX:**
+## Citation
 
-[More Information Needed]
+**BibTeX:**
+```bibtex
+@dataset{fiftyone_gui_grounding_2025,
+  title={FiftyOne GUI Grounding Training Set},
+  author={Sahota, Harpreet},
+  year={2025},
+  publisher={Hugging Face},
+  url={https://huggingface.co/datasets/harpreetsahota/FiftyOne-GUI-Grounding-Train}
+}
+```
 
 **APA:**
+Sahota, H. (2025). FiftyOne GUI Grounding Training Set [Dataset]. Hugging Face. https://huggingface.co/datasets/harpreetsahota/FiftyOne-GUI-Grounding-Train
 
-[More Information Needed]
-
-## Glossary [optional]
-
-<!-- If relevant, include terms and calculations in this section that can help readers understand the dataset or dataset card. -->
-
-[More Information Needed]
-
-## More Information [optional]
-
-[More Information Needed]
-
-## Dataset Card Authors [optional]
+## Dataset Card Authors
 
-[More Information Needed]
+Harpreet Sahota
 
 ## Dataset Card Contact
 
-[More Information Needed]
+For questions about this dataset, please contact the dataset author through the Hugging Face dataset repository.