Automated Detection of Label Errors in Semantic Segmentation Datasets via Deep Learning and Uncertainty Quantification.
Matthias RottmannMarco ReesePublished in: CoRR (2022)
Keyphrases
- deep learning
- automated detection
- semantic segmentation
- weakly supervised
- pascal voc
- superpixels
- bag of words
- object detection
- unsupervised learning
- object recognition
- object class
- object categories
- machine learning
- object classes
- object segmentation
- semi supervised
- conditional random fields
- supervised learning
- relation extraction
- segmentation method
- named entities
- maximum likelihood
- text classification
- topic models
- bounding box
- natural images
- co occurrence
- image features
- viewpoint
- computer vision
- learning algorithm