CS 461 FInal HomeworkDue 12/15 (note change)1. a) What are the three approaches to object recognition that we discussed?b) For each algorithm, state whether or not it can be applied to: i. 2D-2D object recognition ii. 2D-3D object recognition iii. 3D-3D object recognition2. In general, computing 3D pose from 3D data requires computing bothtranslation and rotation. However, we were able to separate it into a problemof computing rotation, and then computing translation. What trick did weuse? If you want, describe it in terms of the basic 3D-3D pose equation qi = Rpi + t. Where else did we use this trick?3. In the paper by David Lowe, he developed an algorithm that can be viewed asstarting with local features that are not highly invariant and progressing tolarger features with more invariance. In this regard:a. What transformations (image changes) are SIFT keys designed to beinvariant to?b. How many sift keys are needed for testing alignment? Under what setof transformations?c. Does the method make use of explicit 3D geometry?Below is a model and some observed data. I’ve supplied the tables for thedistance and angle constraints for model and data. On the next page is a partialinterpretation tree. First, prune any of the existing branches at the second levelthat can be pruned. Finally, add any consistent third level branches to theremaining paths. How many consistent interpretations of this data are there?What are the feature correspondences?anglesF1F2F3F4F5F6F103!/203!/2!!/2F2!/20!/203!/2!F303!/203!/2!!/2F4!/20!/203!/2!F5!!/2!!/203!/2F63!/2!3!/2!!/20anglesf1f2f3f10!/2!f23!/20!/2f3!3!/20distancef1f2f3f1[0,1][1,5][1,2]f2[1,5][0,1][1,5]f3[1,2][1,5][0,1]distanceF1F2F3F4F5F6F1[0,1][0,5][4,8][5,13][9,13][0,10]F2[0,5][0,4][0,5][1,10][1,10][1,10]F3[4,8][0,5][0,1][0,2][1,5][1,8]F4[5,13][1,10][0,2][0,1][0,5][4,13]F5[9,13][1,10][1,5][0,5][0,4][0,13]F6[0,10][1,10][1,8][4,13][0,13][0,9]F1F2F3F4F5F6f3f2f1ModelDataAngles must beequalObserved distancemust be a subset ofmodel