Recent Advances in Computer Vision

6,904 views

Published on

Published in: Education
2 Comments
8 Likes
Statistics
Notes
No Downloads
Views
Total views
6,904
On SlideShare
0
From Embeds
0
Number of Embeds
39
Actions
Shares
0
Downloads
142
Comments
2
Likes
8
Embeds 0
No embeds

No notes for slide

Recent Advances in Computer Vision

  1. 1. Recent Advances in Computer Vision Ming-Hsuan Yang
  2. 2. Computer vision• Holly grail – tell a story from an image
  3. 3. History• “In the 1960s, almost no one realized that machine vision was difficult.” – David Marr, 1982• Marvin Minsky asked Gerald Jay Sussman to “spend the summer linking a camera to a computer and getting the computer to describe what it saw” – Crevier, 1993• 40+ years later, we are still working on this
  4. 4. 1970s
  5. 5. 1980s
  6. 6. 1990s • Face detection • Particle filter • Pfinder • Normalized cut
  7. 7. 2000s• SIFT – Mosaicing, panorama – Object recognition – Photo tourism, photosynth – Human detection• Adaboost-based face detector
  8. 8. Related topics
  9. 9. Conferences• CVPR – Computer Vision and Pattern Recognition, since 1983 – Annual, held in US• ICCV – International Conference on Computer Vision, since 1987 – Every other year, alternate in 3 continents• ECCV – European Conference on Computer Vision, since 1990 – Every other year, held in Europe
  10. 10. Conferences (cont’d)• ACCV – Asian Conference on Computer Vision• BMVC – British Machine Vision Conference• ICPR – International Conference on Pattern Recognition• SIGGRAPH• NIPS – Neural Information Processing Systems
  11. 11. Conferences (cont’d)• MICCAI – Medical Image Computing and Computer-Assisted Intervention• ISBI – International Symposium on Biomedical Imaging• FG – IEEE Conference on Automatic Face and Gesture Recognition• ICCP, ICDR, ICVS, DAGM, CAIP, MVA, AAAI, IJCAI, ICML, ICRA, ICASSP, ICIP, SPIE, DCC, WACV, 3DPVT, ACM Multimedia, ICME, …
  12. 12. Conference organization• General chairs: administration• Program chairs: handling papers• Area chairs: – Assign reviewers – Read reviews and rebuttals – Consolidation reports – Recommendation• Reviewers• Authors
  13. 13. Review process• Submission• CVPR/ECCV/ICCV – Double blind review – Program chairs: assign papers to area chairs – Area chairs: assign papers to reviewers• Rebuttal
  14. 14. Area chair meetings• 2 day meetings• Several panels• Each paper is reviewed by at least 2 area chairs• Buddy system• Area chair make recommendations• Program chairs make final decisions
  15. 15. Conference acceptance rates• ICCV/CVPR/ECCV: ~ 30%• ACCV (2009): ~ 30%• NIPS: ~ 30%• BMVC: ~ 40%• ICIP: ~ 45%• ICPR: ~ 55%• Disclaimer – low acceptance rate = high quality?
  16. 16. CVPR Submission Oral Overall
  17. 17. ICCV Submission Oral Overall
  18. 18. ECCV Submission Oral Overall
  19. 19. Journals• PAMI – IEEE Transactions on Pattern Analysis and Machine Intelligence, since 1979 (impact factor: 5.96, #1 in all engineering and AI, top-ranked IEEE and CS journal)• IJCV – International Journal on Computer Vision, since 1988 (impact factor: 5.36, #2 in all engineering and AI)• CVIU – Computer Vision and Image Understanding, since 1972 (impact factor: 2.20)
  20. 20. Journals (cont’d)• IVC – Image and Vision Computing• IEEE Transactions on Medical Imaging• TIP – IEEE Transactions on Image Processing• MVA – Machine Vision and Applications• PR – Pattern Recognition• TM – IEEE Transactions on Multimedia• …
  21. 21. PAMI review process• Editor-in-chief (EIC) assigns papers to associate editors (AE)• AE assigns reviewers• First-round review: 3-6 months – Accept as is – Accept with minor revision – Major revision – Resubmit as new – Reject
  22. 22. PAMI review process (cont’d)• Second-round review: 2-4 months – Accept as is – Accept with minor revision – Reject• EIC makes final decision• Overall turn-around time: 6 to 12 months• Rule of thumb: 30% additional work beyond a CVPR/ICCV/ECCV paper
  23. 23. IJCV/CVIU review process• Similar formats• CVIU has roughly the same turn-around time as PAMI• IJCV tends to have longer turn-around time
  24. 24. Journal acceptance rate• PAMI, IJCV: ~ 30%• CVIU: ~ 30%
  25. 25. Tools• Google scholar, citeseer,• h-index• Software: publish or perish• Disclaimer: – h index = significance? – # of citation = significance
  26. 26. How to get your papers rejected?• Refer to Jim Kajia (SIGGRAPH 93 papers chair): How to get your SIGGRAPH paper rejected?• Do not – Pay attention to review process – Put yourself as a reviewer perspective – Put the work in right context – Carry out sufficient amount of experiments – Compare with state-of-the-art algorithms – Pay attention to writing
  27. 27. Pay attention to review process• Learn how others/you can pick apart a paper• Learn from other’s mistakes• Learn how to write good papers• Learn what it takes to get a paper published
  28. 28. Put yourself as reviewer• What are the contributions?• Does it advance the science in the filed?• Why you should accept this paper?• Is this paper a case study?• Is this paper interesting?• What is the audience?• Does anyone care about this work?
  29. 29. Experimental validation• Common data set• Killer data set• Large scale experiment• Evaluation metric
  30. 30. Compare with state of the art• Do your homework• Need to know what is out there• Need to show why one’s method outperforms others, and in what way? – speed? – accuracy? – easy to implement? – general application?
  31. 31. Writing• Clear presentation• Terse• Careful about wording• Make claims with strong evidence
  32. 32. Review form• Summary• Overall Rating – Definite accept, weakly accept, borderline, weakly reject, definite reject• Novelty – Very original, original, minor originality, has been done before• Importance/relevance – Of broad interest, interesting to a subarea, interesting only to a small number of attendees, out of CVPR scope
  33. 33. Review form (cont’d)• Clarity of presentation – Reads very well, is clear enough, difficult to read, unreadable• Technical correctness – Definite correct, probably correct but did not check completely, contains rectifiable errors, has major problems• Experimental validation – Excellent validation or N/A (a theoretical paper), limited but convincing, lacking in some aspects, insufficient validation• Additional comments• Reviewer’s name
  34. 34. Challenging issues• Large scale• Unconstrained• Real-time• Robustness• Recover from failure – graceful dead
  35. 35. Some hot topics• Object recognition, categorization• Internet scale image search• Video search• Human detection• 3D human pose estimation• Computational photography• Scene understanding
  36. 36. Some hot tools• Prior• Context• Sparse representation• Multiple instance learning• Online learning• Convex optimization• Constraint• Hashing
  37. 37. Prior Torralba and Sinha ICCV 01
  38. 38. Prior Heitz and Koller ECCV 08
  39. 39. Prior Jia CVPR 08 He et al. CVPR 09
  40. 40. Scene understanding Leibe et al. CVPR 07
  41. 41. Image search Wu et al. CVPR 09
  42. 42. Computational photography Johnson and Adelson et al. CVPR 09
  43. 43. Computational photography Ahuja et al.

×