Learning Montages of Transformed Latent Images as Representations of Objects That Change in Appearance.
Chris PalBrendan J. FreyNebojsa JojicPublished in: ECCV (4) (2002)
Keyphrases
- visual appearance
- image data
- image classification
- bounding box
- geometric information
- image features
- viewing conditions
- camera viewpoint
- visual patterns
- image database
- image registration
- input image
- viewing angle
- real world objects
- learning algorithm
- complex scenes
- image regions
- d objects
- visual objects
- appearance features
- outdoor scenes
- multiple objects
- target object
- feature points
- object recognition
- object detectors
- object models
- latent variables
- static images
- test images
- appearance model
- keypoints
- visual features
- image retrieval
- surface properties
- three dimensional