Introduction:
Welcome to the page for subjective test of Apidis summarization system. There are three major concepts in our summarization, i.e, content pattern, continuity, and redundancy. Content pattern decides whether the summary focus on telling the game or on showing emotional highlights, which is achieved by tuning the ratio between game relevance and emotional level of clips, denoted as G/E.
Continuity means that more successive clips are selected for a local story to tell an event more fluently, which is controlled by C. With our limited and brief annotation of events and our way of assigning benefits considering the role of difference view types in producing sport videos, continuity also favors completeness of the story by providing longer and continuous local stories in a simple but natural way.
Redundancy controls the frequency of replays and close views in generated summaries, denoted as R. We use x-Medium, x-Low, and x-High to denote those cases that we use a medium, low and high value for parameter x, respectively.
Please view the video and answer the questions to help us finish a subjective test.
* Note that the TV rights of the soccer video used in the following tests are belonging to MEDIAPRO, MP, Spain. All videos in this page are only used for research purpose within the APIDIS project through online streaming.
Downloading or redistributing the videos are not allowed.
|
|
1. May we know your name?
|
|
|
2. Do you like football?(Required)
|
Yes
No
|
| |
| Test 1: Results under different ratios between game relevance and emotional level [C-Medium, R-Medium] |
|
|
| Six Minutes, G/E-High, Emphasize on Game Evolving |
Six Minutes, G/E-Low, Emphasize on Emotional Moment |
|
3. Did you feel the difference by changing the content pattern? (Required)
|
Yes
No
|
|
4. Do you perfer one of the videos in this test?(Required)
|
Left
Right
None
|
|
5. Did you have any comment, or notice some artifacts that must be improved from this test?
|
|
|
| |
| Test 2: Results under different redundancy penalty [G/E-Medium, C-Medium] |
|
|
|
| Six Minutes, R-Low, Less Replays/Close Views, More Events |
Six Minutes, R-High, More Replays/Close Views, Less Events |
|
6. Did you feel the difference by changing the content redundancy? (Required)
|
Yes
No
|
|
7. Do you perfer one of the videos in this test? (Required)
|
Left
Right
None
|
|
8. Did you have any comment, or notice some artifacts that must be improved from this test?
|
|
|
| |
| Test 3: Results under different continuity gain [G/E-Medium, R-Medium] |
|
|
|
| Six Minutes, C-Low, More but Shorter Local Stories of Events |
Six Minutes, C-High, Less but Longer Local Stories of Events |
|
9. Did you feel the difference by changing the content continuity? (Required)
|
Yes
No
|
|
10. Do you perfer one of the videos in this test? (Required)
|
Left
Right
None
|
|
11. Did you have any comment, or notice some artifacts that must be improved from this test?
|
|
|
| |
| Test 4: Results under different length of summary [G/E-Medium, C-Medium, R-Medium] |
|
|
|
| One Minute |
Three Minutes |
|
|
|
| Six Minutes |
Twelve Minutes |
|
12. Did the way of summary expanding under different lengths meet your expectation? (Required)
|
Yes
No
|
|
13. Did you see some visual artifacts that must be improved from this test?
|
|
|
| |
|
14. Do you have any other comments that you would like to say?
|
|
|
|
|