9 distinct classes of noisy images organized for classification tasks. The collection provides visual data specifically curated to include noise artifacts across all categories.
Use Cases
- Train a noise-tolerant image classifier using the 9 provided class labels
- Evaluate the impact of image noise on standard computer vision architectures using the noisy image samples
- Test image preprocessing and denoising filters on the noisy image samples before classification
Strengths
- Includes 9 distinct image classes for categorical labeling
- Consists of image data specifically characterized by noise artifacts
- Provides a labeled set of noisy images for supervised learning