dynamicslide reference based interaction techniques for
play

DynamicSlide: Reference-based Interaction Techniques for - PowerPoint PPT Presentation

DynamicSlide: Reference-based Interaction Techniques for Slide-based Lecture Videos Hyeungshik Jung, KAIST Hijung Valentina Shin, Adobe Research Juho Kim, KAIST MOOC and video 2


  1. DynamicSlide: Reference-based Interaction Techniques for Slide-based Lecture Videos Hyeungshik Jung, KAIST Hijung Valentina Shin, Adobe Research Juho Kim, KAIST

  2. MOOC and video 2 https://www.class-central.com/report/moocs-stats-and-trends-2017/

  3. Slides are widely used in lecture videos 3

  4. Decomposing Slide-based Video Frame Audio 4

  5. Decomposing Slide-based Video Slide Audio 5

  6. Decomposing Slide-based Video Slide Audio 6

  7. Decomposing Slide-based Video timeframe “Hash function” “Hash function” Slide “Takes any string as input” Image1 “Fixed-size output (we’ll use 256 bits)” Image1 Audio 7

  8. Decomposing Slide-based Video timeframe “Hash function” “Hash function” Frame “Takes any string as input” Image1 “Fixed-size output (we’ll use 256 bits)” Image1 Audio 8

  9. Object-oriented tools for both learners & instructors For Learners (Player) For Instructors (Lightweight editor) 9

  10. Object-oriented tools for both learners & instructors For Learners (Player) For Instructors (Lightweight editor) 10

  11. Link between slide and narration 11

  12. Link between slide and narration visual information hierarchy 12

  13. Link between slide and narration auditory visual detailed explanation information hierarchy 13

  14. Link between slide and narration How can we find leverage links between slides and narration for watching video? 14

  15. Interaction techniques powered by links between slides and narration 1. Automatic Highlighting 2. Item-based Navigation 3. In-video Bookmarking 15

  16. Interaction techniques powered by links between slides and narration 1. Automatic Highlighting 2. Item-based Navigation 3. In-video bookmarking 16

  17. Computational Pipeline 17

  18. Computational Pipeline Pixel difference Text difference 18

  19. Computational Pipeline Heuristic 19

  20. Computational Pipeline 20

  21. Computational Pipeline 21

  22. Computational Pipeline 22

  23. User Study (n=12) Baseline DynamicSlide 1. Navigation -> Searching relevant part of the question 2. Cognitive load -> NASA-TLX 23

  24. User Study (n=12) Navigation task Cognitive Load 24

  25. Discussion ● How to robustly find link between slides and narrations? ● Do links commonly exist in lecture videos? ● The effect of finding and leveraging wrong link 25

  26. Object-oriented tools for both learners & instructors For Students (Player) For Instructors (Lightweight editor) 26

  27. Lifespan of a lecture t 27

  28. Lifespan of a lecture ● Update the content of slide ● Try different versions of explanation t 28

  29. Lifespan of a lecture ● Update the content of slide ● Try different versions of explanation t ● Hard to update the published video 29

  30. Lifespan of a lecture ● Update the content of slide ● Try different versions of explanation t ● Hard to update the published video How can we build a tool to help instructors directly update their video without having to patch and re-publish? 30

  31. Formative study Interviews with 7 Instructors who made lecture video (Professors + Part time) 3 Video editors (Content team at the university) 31

  32. Formative study 1. Instructors want to edit the videos before and after publications Interviews with 2. Editing cost is prohibitive 3. Editing often involves substituting the original 7 Instructors who made lecture video (slide) contents (Professor + Part time) 3 Video editors (Content team at the university) 32

  33. Formative study 1. Instructors want to edit the videos before and after publications Interviews with 2. Editing cost is prohibitive 3. Editing often involves substituting the original 7 Instructors who made lecture video (slide) contents (Professor + Part time) 3 Video editors Before ● (Content team at the university) Add supplementary content ● Highlight the text ● Cut part with mistakes After ● Fix typo ● Fix figure 33

  34. Formative study 1. Instructors want to edit the videos before and after publications Interviews with 2. Editing cost is prohibitive 3. Editing often involves substituting the original 7 Instructors who made lecture video (slide) contents (Professor + Part time) 3 Video editors Before ● (Content team at the university) Add supplementary content ● Highlight the text ● Cut part with mistakes After ● Fix typo ● Fix figure 34

  35. Formative study 1. Instructors want to edit the videos before and after publications Interviews with 2. Editing cost is prohibitive 3. Editing often involves substituting the original 7 Instructors who made lecture video (slide) contents (Professor + Part time) 3 Video editors (Content team at the university) 35

  36. Formative study 1. Instructors want to edit the videos before and after publications Interviews with 2. Editing cost is prohibitive 3. Editing often involves substituting the original 7 Instructors who made lecture video (slide) contents (Professor + Part time) Changing text (ABC -> CBA) 3 Video editors (Content team at the university) ABC 36

  37. Formative study 1. Instructors want to edit the videos before and after publications Interviews with 2. Editing cost is prohibitive 3. Editing often involves substituting the original 7 Instructors who made lecture video (slide) contents (Professor + Part time) Changing text (ABC -> CBA) 3 Video editors (Content team at the university) cba ABC Create patch 37

  38. Formative study 1. Instructors want to edit the videos before and after publications Interviews with 2. Editing cost is prohibitive 3. Editing often involves substituting the original 7 Instructors who made lecture video (slide) contents (Professor + Part time) Changing text (ABC -> CBA) 3 Video editors (Content team at the university) CBA ABC Create patch 38

  39. Formative study 1. Instructors want to edit the videos before and after publications Interviews with 2. Editing cost is prohibitive 3. Editing often involves substituting the original 7 Instructors who made lecture video (slide) contents (Professor + Part time) Changing text (ABC -> CBA) 3 Video editors (Content team at the university) ABC CBA Overwrite + Adjust timing 39

  40. Design Goals 1. Provide an intuitive interface for instructors to directly edit their lecture videos 2. Enable users to update elements of the lecture slide within the video 3. Facilitate synchronization of visual and audio events in the video 40

  41. DynamicLecture 41

  42. DynamicLecture 42

  43. DynamicLecture 43

  44. DynamicLecture 44

  45. DynamicLecture 45

  46. DynamicLecture 46

  47. DynamicLecture 47

  48. Evaluation 1. Instructors review their lecture videos, then list up improvement items 2. Inststructors fix their video using the system 48

  49. Result 1. Instructors review their lecture videos, then list up improvement items 2. Inststructors fix their video using the system 49

  50. Conclusion For Students (Player) 1. Parsing videos into objects can improve the way we interact, consume, & create lecture videos. 2. We need to explore novel object-based video interaction techniques. For Instructors ( e.g., participatory video improvement by learners) (Lightweight editor) 50

Recommend


More recommend