Aurora Image Classification Robust to Noise Labels.
Takeru EndoMitsuharu MatsumotoPublished in: IECON (2021)
Keyphrases
- image classification
- multi label
- image noise
- salt pepper
- feature extraction
- image representation
- pairwise
- label noise
- random noise
- noisy environments
- computer vision
- imaging artifacts
- registration errors
- noisy data
- missing data
- visual features
- noise level
- geometric distortions
- robust statistical
- weak edges
- greater robustness
- image features
- data sets