Tools
Sorted by:
Try your query at:
 |
 |
 |
 |
 |
 |
Results 1 - 10
of
30,040
Table V. Message priority confusion matrices as determined by 16 subjects who were initially presented with the automatically produced text summaries and later with the original audio of 15 messages. Each column corre- sponds to the percentage of messages classi ed as low/medium/high priority based on text summaries and each row gives the percentage of messages classi ed as low/medium/high priority based on the original audio. The degree of consistency across judgements can be determined by summing up the diagonal values.
in Automatic
Table 5.1 gives the CHR rules automatically produced by our library from the JSL specifi- cations given in the example 6.
1725
Table 5.1 gives the CHR rules automatically produced by our library from the JSL speci - cations given in the example 6.
1725
Table 3 compares the automatically produced abstracts with the man-made trailers (where available). Although some scenes are obviously identical, there is still a large amount of di erence, reasons having been mentioned above. The di erence when viewing the abstracts may not be as big, because there are always similar scenes in a video from which both automatic and manual abstracter have to choose, and they usually do not decide on the same scene.
"... In PAGE 11: ... Table3 : Comparison of abstracts and trailers: number of identical scenes... ..."
Table 9: Characterization of query expansion used in best automatic ad hoc runs.
1998
"... In PAGE 13: ... The automatically-produced UMass phrase list was new for TREC-6, the Cornell list was basically unchanged from early TRECs, and the BBN list was based on a new bigram model. Table9 shows characteristics of the expansion tools used in these systems. The second column gives the... ..."
Cited by 65
Table 9: Characterization of query expansion used in best automatic ad hoc runs. Organization Expansion/Feedback Top Docs/Terms added Disks used Comments
1998
"... In PAGE 13: ... The automatically-produced UMass phrase list was new for TREC-6, the Cornell list was basically unchanged from early TRECs, and the BBN list was based on a new bigram model. Table9 shows characteristics of the expansion tools used in these systems. The second column gives the... ..."
Cited by 65
Table 3.2: Summary of experimental results: Optimal is numerically estimated. GFI is the performance of the new algorithm after 200 iterations using automatic range expansion when it produces improved results (automatic range expansion made no signi cant improvements for ranges less than 10000). RFI is from traditional table lookup with random trials. The last column indicates how many trials RFI needs to equal GFI apos;s performance.
Table 1: Overview of analyzed videos and their original trailers
1996
"... In PAGE 11: ... Thus wewere able to compare the man-made abstracts with our automatically produced ones. Table1 gives an overview of the seven videos whichwe analyzed.... ..."
Cited by 43
Table 6: CUED RT-03s segmentation: Segment distribution over gender and bandwidth for each show in bneval03
2004
"... In PAGE 16: ... This gave automatically produced segment labels which contained a start and end time, a gender (male or female) and a bandwidth (wideband or narrowband). The breakdown of the segment distribution by show is given in Table6 .... ..."
Cited by 4
Table 6: CUED RT-03s segmentation: Segment distribution over gender and bandwidth for each show in bneval03
2004
"... In PAGE 17: ... This gave automatically produced segment labels which contained a start and end time, a gender (male or female) and a bandwidth (wideband or narrowband). The breakdown of the segment distribution by show is given in Table6 .... ..."
Cited by 4
Results 1 - 10
of
30,040
