Grocery Detection Dataset - Model Overview
Download Dataset
You can find the dataset here: Download Dataset
1. Dataset Overview
The Grocery Detection Dataset provides comprehensive insights into grocery item recognition for machine learning applications.
Total Dataset Size
24,938 unique images
Annotation Format
Annotated in YOLOv11 format
Pre-processing Applied
- Auto-orientation of pixel data (with EXIF-orientation stripping)
- Resize to 640x640 (Stretch)
Data Augmentation
- 50% probability of horizontal flip
- 50% probability of vertical flip
- Equal probability of one of the following 90-degree rotations: none, clockwise, counter-clockwise, upside-down
- Random rotation of between -45 and +45 degrees
- Random shear of between -45° to +45° horizontally and -45° to +45° vertically
- Random brightness adjustment of between -48% and +48%
- Random exposure adjustment of between -12% and +12%
2. Dataset Distribution Visualization
Dataset Distribution by Set Type
Sample Count Across Dataset Splits
3. Data Split Analysis
Comprehensive breakdown of dataset distribution for model training:
| Set Type | Sample Count | Percentage | Metadata Representation |
|---|---|---|---|
| Training Set | 17,400 | 70% | Primary model training images |
| Validation Set | 4,987 | 20% | Model performance validation images |
| Test Set | 2,551 | 10% | Final model evaluation images |
4. Dataset Collection Challenges
Key challenges encountered during dataset collection:
- Varied Product Formats: Diverse grocery packaging
- Labeling Variability: Inconsistent product descriptions and categorizations
- Complex Backgrounds: Images captured in varied lighting and environments
- Data Augmentation Necessity: Enhancing model robustness through augmentation
5. Conclusion
The Grocery Detection Dataset offers a robust resource for training models to identify grocery items in various conditions, facilitating advancements in automated product recognition.