Recognition using Regions

1. Recognition using Regions Ghunhui Gu, Joseph J. Lim, Pablo Arbeláez, Jitendra Malik University of California at Berkeley Berkeley, CA 94720

2. OUTLINE Introduction Approach Experimental Results Conclusion

3. Introduction • Early work in the late 90s , the domain strategy for object detection in a scene has been multi-scale scanning : is there an instance of object category C in the window?

4. It differs significantly from the nature of human visual detection So, • This paper focus on using regions , which have some properties: • They encode shape and scale information of objects naturally • They specify the domains on which to compute various features, without being affected by clutter from outside the region (background) • But its not popular as features due to their sensitivity to segmentation error

5. Approach Overview the method Framwork for Region weighting Main recognition algorithm (1)Voting (2)Verification (3)Segmentation

6. The “bag of regions” representation of a mug example All node generated by[2] [2] P. Arbel´aez, M. Maire, C. Fowlkes, and J. Malik. From contours to regions: An empirical evaluation. In CVPR, 2009.

7. Region cues: Describe a region by subdividing evenly its bounding box int an n x n grid • Contour shape, given by the histogram of oriented responses of the contour detector gPb [22] • Edge shape, where orientation is given by local image gradient (by convolution) • Color, represented by the L*, a and b histograms in the CIELAB color space http://en.wikipedia.org/wiki/Lab_color_space • Texture, described by texton histograms

8. The “contour shape” region descriptor (a)Original image, (b) A region from the image, (c) gPb[22]Representation of the region in (b), (d) Our contour shape descriptor based on (c) [22] M. Maire, P. Arbel´aez, C. Fowlkes, and M. Malik. Using contours to detect and localize junctions in natural images. In CVPR, 2008.

9. Discriminative Weight Learning I and J are objects of same category, but K is an object of a different category

11. Discriminative Weight Learning

12. The pipeline of object recognition algorithm Voting , Verification, Segmentation three stage

13. Voting stage This transformation provides not only position but also scale estimation of the object. It also allows for aspect ratio deformation of bounding boxes.

14. Given a query image and an object category, is to generate hypotheses of bounding boxes and support of objects of that category in the image Voting Vote of bounding box of the object(Transformation function ) Vote score Transformation function model they use

15. Verification The verification score The average of the probabilities The overall detection score -- Product of the two score

16. Segmentation To recover the complete object support from one of its parts Green for object and Red for background

17. Experimental Results Data base: 1. ETHZ shape 2. Caltech-101

18. Detection performance

19. ETHZ shape • Region tree : on average ~ 100 regions per image • Color and texture are not very useful in this data base Choose the functions in Eqn.11 as: Split the entire set in to half training and half test for each category

20. ETHZ shapes

21. Caltech 101 Randomly pick 5, 15 or 30 images for training and up to 15 images in disjoint set for test Geometric blur[4]

22. Caltech 101

23. conclusion • Presented a unified framework for object detection, segmentation, and classification using regions. • (1)Cue combination significantly boosts recognition performance • (2)Reduces the number of candidate bounding box by order of magnitude over standard sliding window scheme due to robust estimation of object scales from region matching