Object Recognition

Object Recognition

UCF CRCV via YouTube Direct link

Selective search to get high recall

43 of 47

43 of 47

Selective search to get high recall

Class Central Classrooms beta

YouTube videos curated by Class Central.

Classroom Contents

Object Recognition

Automatically move to the next video in the Classroom when playback concludes

  1. 1 Object recognition
  2. 2 Bag of Words
  3. 3 Bags of Words
  4. 4 Capture the pattern in patch
  5. 5 The dimensionality
  6. 6 Sample many patches
  7. 7 Two types of sampling
  8. 8 Sample many images
  9. 9 Include all relevant variations
  10. 10 Form a dictionary of words
  11. 11 Ideally words cover similar patches
  12. 12 Count words per image
  13. 13 Codeboods
  14. 14 Learn histogram similarity
  15. 15 Similarity between two histograms
  16. 16 Classify unknown image
  17. 17 Concept-specific codebooks
  18. 18 Conclusion on concept-codebooks
  19. 19 Soft word assignment
  20. 20 Fisher vector
  21. 21 Conclusion on words
  22. 22 Codebook synonyms
  23. 23 How close are synonyms?
  24. 24 90% removed, same result
  25. 25 Visual synonym examples
  26. 26 Conclusion on visual synonyms
  27. 27 Convex reduced codebooks
  28. 28 Conclusion convex reduced
  29. 29 The where and what
  30. 30 What makes a boat a boat?
  31. 31 What is the object in the middle?
  32. 32 Where is evidence?
  33. 33 Where is evidence for an object?
  34. 34 The visual extent of an object
  35. 35 Context dominance
  36. 36 Object dominance
  37. 37 Object detail dominance
  38. 38 Pyramids: simple compositional
  39. 39 Exhaustive search
  40. 40 The need for high recall
  41. 41 The need for hierarchy
  42. 42 Selective search example
  43. 43 Selective search to get high recall
  44. 44 Average best overlap -88%
  45. 45 Classification with selective search
  46. 46 Conclusion on location
  47. 47 Two concepts in interaction

Never Stop Learning.

Get personalized course recommendations, track subjects and courses with reminders, and more.

Someone learning on their laptop while sitting on the floor.