Dataset Guide
You need to be a Pro member to access the full content of this.
- 1. Download real dataset
- Download an example dataset
- Unzip the dataset
- 2. Verify dataset visually
- 1. Visualize the JSON files
- 2. What we can see
- 3. Refine the annotation
- 3. Verify dataset statistically
- 1. Show stats
- 2. Find overall insights
- 3. Find insights on each annotated label
- 4. Find insights on each annotated file
- 5. Per-file & label stats
- 4. Additional data collection
- 1. Compatibility
- 2. Progressiveness
- 3. Extendability
- 5. Export dataset
- Convert dataset to a common format (masks)
- Use exported files from Python
- Bonus: PyTorch dataset class
- Write PyTorch dataset class
- Verify the dataset class
- Train your model