Extracting templates for
Natural Scene Classification

Content Based Image Retrieval

Difficulty

	Variations in color, texture, spatial properties

Configural Recognition
(Lipson 1996)

	Predefined templates :
		relative spatial and photometric properties between image regions
		low resolution
		global configuration of image regions

Pre-defined Template
(Lipson 1996)

Template Extraction

Example:
Extracting the Snowy Mountain Template

Multiple-Instance learning for Natural Scene Classification

Slide 9

Multiple Instance Learning

	Regular learning: learn concept form leabeled examples
	MI learning: learn concepts from labeled bags
	a bag is a collection of examples
	positive bag has at least one positive example
	negative bag has only negative examples

Natural Scene Classification

	Give me more images like this
	Images are inherently ambiguous
	Be explicit about the ambiguity: an image is a bag and each instance is something that possible represents the image

Slide 12

Slide 13

HYPOTHESIS  CLASSES

SNAPSHOT  OF  SYSTEM

Slide 16

Slide 17

Slide 18

GRAPHS
Comparison with hand-crafted templates

Slide 20

Precision curve for different training schemes
averaged over concepts and hypotheses.

Dataset = 400 images

Slide 22

Car Detection

Vehicle detection: Example

Some Results

Results on video sequence