1. Pyramids of Features For Categorization Greg Griffin and Will Coulter (see Lazebnik et al., CVPR 2006, too)

2. Intuition: Approximates optimal partial matching U Adapted from http://people.csail.mit.edu/kgrauman/slides/pyr_match_iccv2005.ppt

3. Intuition [cont’d]: Combine bags of features with spatial information

4. Example Pyramid Comparison

5. Disadvantages Same objects in different quadrants Objects sliced by bins

6. Possible Solutions Flipping / rotating image Sliding / shuffling histogram bins

7. Possible Solutions [cont’d] Split histogram in powers of three

8. Implementation Overview Image and Feature Extraction (SIFT on regular grid) ↓ Vocabulary Translation (200 words (k-means)) ↓ Histogram Generation (flips, slides, arbitrary mixing) ↓ Matching (full or partial pyramid) ↓ Decision (best match, voting, SVM)

9. Sanity Check 1 Graphical mini-confusion matrix (and rot. invariance)

10. Sanity Check 2

11. Scene Database

12. 81.1% vs 67.4% (100)

13. Caltech 101 64.6% vs. 33.1% (30, 16)

14. Caltech 101 minaret windsor chair joshua tree okapi cougar body beaver crocodile ant 97.6%, 86.7% 94.6%, 80.0% 87.9%, 40.0% 87.8%, 33.3% 27.6%, 6.7% 27.5%, 13.3% 25.0%, 13.3% 25.0%, 0.0% (30, 16)

15. Caltech 101

16. Work in Progress 256 Performance –64 times more work scene database –6.4 times more work than 101 SVM –one-vs-all weighting issues –speed it up? –improve performance Improvements –Flip, Slide, Arbitrary Powers-of-3 histogram bins

17. Open Questions Performance of arbitrary match bins –Try random sampling? –Allow multiple best matches? Chess/pattern example Grid example Optimal kernel level weights

19. Implementation Details [block diagram] Images (288^2,b&w,squished) and feature extraction (-weak,-pca,+sift) Vocab generation (200 words, 20,000- small) Histogram(fliplr,flipud,slide,arbitrary,bag of features) match (full&partial pyramids) Decision (best match,voting,SVM)