A Centroid Loss for Weakly Supervised Semantic Segmentation in Quality Control and Inspection Application.
Kai YaoAlberto OrtizFrancisco Bonnín-PascualPublished in: CoRR (2020)
Keyphrases
- quality control
- weakly supervised
- semantic segmentation
- superpixels
- object class
- relation extraction
- scene classification
- topic models
- information extraction
- conditional random fields
- named entities
- object classes
- object categories
- object recognition
- object segmentation
- computer vision
- segmentation algorithm
- pascal voc
- object detection