UnAMT: unsupervised adaptive matting tool for large-scale object collections.
Jaehwan KimJongyoul ParkKyoung ParkPublished in: SIGGRAPH Posters (2015)
Keyphrases
- d objects
- high quality
- data driven
- digital objects
- machine learning
- information retrieval
- object model
- digital libraries
- complex objects
- unsupervised learning
- object tracking
- neural network
- small scale
- data objects
- lighting conditions
- metadata
- data sets
- document collections
- semi supervised
- spatial relationships
- moving objects
- object segmentation
- foreground and background