- STSIM vector extraction
- Compute covariance matrix
- Approximate Cross Validation
- Compute cluster center or set example for retrieval
- Measure Precision@1 and RoC
- Global or Intraclass covariance matrix calculation
- Color or Grayscale features
- CuReT: Fewer classes, many examples per class
- Represent class with many exemplars
- Tests under variation of lighting and viewing conditions
- Identical: Same texture, perceptually identical
- ViSiProg: Different classes; Perceptually similar
- Distortions: Single image with applied distortion translation, rotations, scaling, etc.
- Previous work would have made assumptions among feature correlation
- Similarity is order invariant
- 1-,2-Color improves accuracym; Minimal gains for 4-color