telefonica research trecvid 2011
play

Telefonica Research @ Trecvid 2011 Xavier Anguera, Daru Xu - PowerPoint PPT Presentation

Telefonica Research @ Trecvid 2011 Xavier Anguera, Daru Xu 1 and Tomasz Adamek (With the collaboraBon of Juan Manuel Barrios, Prisma Group) 1 Daru Xu


  1. Telefonica ¡Research ¡ @ ¡Trecvid ¡2011 ¡ Xavier ¡Anguera, ¡Daru ¡Xu 1 ¡and ¡Tomasz ¡ Adamek ¡ (With ¡the ¡collaboraBon ¡of ¡Juan ¡Manuel ¡ Barrios, ¡Prisma ¡Group) ¡ 1 Daru ¡Xu ¡is ¡a ¡graduate ¡student ¡at ¡the ¡Ming-­‑Hsieh ¡Department ¡of ¡Electrical ¡Engineering, ¡ University ¡of ¡Southern ¡California, ¡USA ¡ ¡ ¡

  2. Outline ¡of ¡the ¡talk ¡ • Telefonica ¡2011 ¡Video-­‑copy ¡detecBon ¡system ¡ – Overall ¡system ¡ – Video-­‑copy ¡detecBon ¡ – Audio-­‑copy ¡detecBon ¡ – Fusion ¡algorithm ¡ – Results ¡ • MulB-­‑systems ¡fusion ¡experiment ¡

  3. MulBmodal ¡Video-­‑copy ¡detecBon ¡

  4. Video-­‑based ¡System ¡ DART* ¡local ¡features ¡extracBon ¡ Video ¡query ¡ Inserted ¡staBc ¡ Key-­‑frame ¡ DART* ¡ text ¡& ¡banners ¡ extracBon ¡ extracBon ¡ filtering ¡ SubBtle ¡ Temporal ¡stability ¡& ¡ Key-­‑frame ¡ filtering ¡ scale ¡filtering ¡ matching ¡ ¡ Ref. ¡Video ¡ indexing ¡ info. ¡ Matched ¡video ¡ segments ¡ Differences ¡from ¡last ¡year: ¡ Temporal ¡ consistency ¡ ¡ • So]ware ¡refactoring ¡ post-­‑processing ¡ • EliminaBon ¡of ¡temporary ¡files ¡ * ¡D. ¡Marimon, ¡A. ¡Bonnin, ¡T. ¡Adamek, ¡and ¡R ¡.Gimeno, ¡“DARTs:Efficient ¡scale-­‑space ¡ extracBon ¡of ¡daisy ¡key-­‑points”, ¡CVPR ¡2009. ¡

  5. Audio-­‑based ¡System ¡

  6. MASK ¡fingerprint ¡extracBon ¡(I) ¡ 1) ¡Audio ¡track ¡extracBon ¡using ¡FFMPEG ¡

  7. AcousBc ¡fingerprint ¡extracBon ¡(I) ¡ 1) ¡Audio ¡track ¡extracBon ¡using ¡FFMPEG ¡ 10ms, ¡100ms ¡window ¡ 2) ¡FFT, ¡bandwidth ¡ limited ¡to ¡ 300-­‑3KHz ¡ 32 ¡MEL-­‑spectrum ¡bands ¡

  8. AcousBc ¡fingerprint ¡extracBon ¡(I) ¡ 1) ¡Audio ¡track ¡extracBon ¡using ¡FFMPEG ¡ 10ms, ¡100ms ¡window ¡ 2) ¡FFT, ¡bandwidth ¡ limited ¡to ¡ 3) ¡Find ¡spectrogram ¡ 300-­‑3KHz ¡ peaks. ¡ 32 ¡MEL-­‑spectrum ¡bands ¡

  9. AcousBc ¡fingerprint ¡extracBon ¡(II) ¡ 4) ¡Apply ¡a ¡mask ¡in ¡each ¡ maxima ¡locaBon ¡

  10. AcousBc ¡fingerprint ¡extracBon ¡(II) ¡ 5) ¡Construct ¡the ¡fingerprint ¡

  11. MulBmodal ¡Fusion ¡Algorithm ¡ • Fusion ¡of ¡different ¡modaliBes ¡at ¡decision ¡level ¡ – AgnosBc ¡of ¡internal ¡system’s ¡behaviors ¡ • No ¡limit ¡on ¡the ¡number ¡of ¡systems ¡to ¡be ¡combined ¡ – provided ¡each ¡system ¡is ¡bejer ¡than ¡random ¡ • To ¡work ¡opBmally ¡it ¡needs ¡N-­‑best ¡matches ¡from ¡each ¡ system. ¡It ¡returns ¡the ¡best ¡fused ¡matches ¡(N=20) ¡ – Makes ¡use ¡of ¡the ¡individual ¡scores ¡and ¡the ¡rank ¡within ¡ each ¡modality. ¡ Paper ¡on ¡ACM ¡MM ¡2011: ¡“MulBmodal ¡Fusion ¡for ¡Video ¡Copy ¡DetecBon”, ¡Xavier ¡Anguera, ¡ ¡ Juan ¡Manuel ¡Barrios, ¡Tomasz ¡Adamek ¡and ¡Nuria ¡Oliver ¡

  12. Data ¡preprocessing ¡ Audio ¡scores ¡histogram ¡ Local ¡video ¡scores ¡histogram ¡ Global ¡video ¡scores ¡histogram ¡

  13. N-­‑best ¡flooring ¡and ¡L1 ¡NormalizaBon ¡(I) ¡ MScores ¡ MScores ¡ 0.7 ¡ 0.3 ¡ L1 ¡normalizaBon ¡ MScores ¡ MScores ¡ MScore i MScore i = 0.6 ¡ Nbest ! MScore j j = 1 0.3 ¡

  14. N-­‑best ¡flooring ¡and ¡L1 ¡NormalizaBon ¡(II) ¡ MScores ¡ MScores ¡ 0.7 ¡ 0.3 ¡ N-­‑best ¡flooring ¡and ¡ ¡ L1 ¡normalizaBon ¡ MScores ¡ MScores ¡ MScore i MScore i = Nbest ! MScore j j = 1 0.3 ¡ 0.2 ¡ N-­‑best ¡Flooring ¡

  15. Overlapping ¡Segments ¡Merge ¡ Q Q B r E r Segment ¡Q ¡ + ¡ R R B r E r min { E Q k ( r ) } − max { B Q Segment ¡R ¡ k ( r ) , E R k ( r ) , B R k ( r ) } = ¡ > 0 . 5 E r max { E Q k ( r ) } − min { B Q B r k ( r ) , E R k ( r ) , B R k ( r ) } Merged ¡segment ¡ Examples: ¡ MuBmodal ¡ Missing ¡ ¡ Non-­‑overlapping ¡ overlap ¡ modality ¡ modaliBes ¡

  16. Output ¡score ¡computaBon ¡ Number ¡of ¡matches ¡ Rank ¡[1 ¡to ¡N k ] ¡ ResulBng ¡score ¡ Normalized ¡ for ¡fused ¡match ¡ matching ¡score ¡ at ¡rank ¡r ¡ A-­‑priori ¡weight ¡for ¡ Best ¡normalized ¡ each ¡modality ¡ matching ¡score ¡for ¡ each ¡modality ¡

  17. Official ¡evaluaBon ¡results ¡ OpBmum ¡scores, ¡balanced ¡profile: ¡ Profile ¡ Min ¡ FA ¡count ¡ Miss ¡ True ¡ Opt ¡F1 ¡ NDCR ¡ count ¡ posi:ves ¡ score ¡ Audio ¡system ¡ BALANCED ¡ 0.662 ¡ 0.66 ¡ 54.75 ¡ 54.78 ¡ 0.729 ¡ MulBmodal ¡ BALANCED ¡ 0.610 ¡ 0.80 ¡ 11.73 ¡ 63.69 ¡ 0.947 ¡ Joint ¡ BALANCED ¡ 0.268 ¡ 0.23 ¡ 4.71 ¡ 101.4 ¡ 0.957 ¡ Choosing ¡only ¡1 st -­‑best ¡results: ¡ ¡ Profile ¡ Min ¡ FA ¡count ¡ Miss ¡ True ¡ Opt ¡F1 ¡ NDCR ¡ count ¡ posi:ves ¡ score ¡ Audio ¡system ¡ BALANCED ¡ 0.477 ¡ 0.14 ¡ 55.89 ¡ 72.05 ¡ 0.712 ¡

  18. MulB-­‑systems ¡fusion ¡experiment ¡ • We ¡tested ¡the ¡fusion ¡algorithm ¡with ¡many ¡ system ¡outputs ¡ • We ¡asked ¡parBcipants ¡in ¡TRECVID ¡2011 ¡for ¡their ¡ submijed ¡runs ¡ – 10 ¡teams ¡contributed ¡their ¡results: ¡PKU-­‑IDM, ¡CRIM, ¡ INRIA-­‑TEXMEX/LEAR, ¡FT, ¡prisma, ¡ATTLabs, ¡kddi, ¡iupr-­‑ dti, ¡brno, ¡Telefonica ¡Research ¡ – I ¡used ¡the ¡“Balanced” ¡runs: ¡17 ¡runs ¡

  19. Status ¡of ¡the ¡runs ¡ • The ¡fusion ¡algorithm ¡works ¡opBmally ¡when ¡ Nbest ¡results ¡are ¡available ¡for ¡each ¡fused ¡ output. ¡ – Results ¡for ¡the ¡used ¡systems ¡had ¡(many ¡Bmes) ¡ only ¡1best ¡results, ¡resulBng ¡subopBmal ¡for ¡the ¡ fusion. ¡

  20. Individual ¡results ¡(Min ¡NDCR) ¡ 0.991 ¡ 1 ¡ 0.9 ¡ 0.8 ¡ Min_NDCR ¡ 0.7 ¡ 0.6 ¡ 0.5 ¡ 0.4 ¡ 0.3 ¡ 0.2 ¡ 0.053 ¡ 0.1 ¡ 0 ¡ 1 ¡ 2 ¡ 3 ¡ 4 ¡ 5 ¡ 6 ¡ 7 ¡ 8 ¡ 9 ¡ 10 ¡ 11 ¡ 12 ¡ 13 ¡ 14 ¡ 15 ¡ 16 ¡ 17 ¡ • Labeled ¡from ¡1 ¡to ¡17, ¡to ¡anonymize ¡them. ¡

  21. Individual ¡results ¡(opBmum ¡F1) ¡ 1 ¡ 0.95 ¡ 0.9 ¡ Min_NDCR ¡ 0.85 ¡ 0.8 ¡ 0.75 ¡ 0.7 ¡ 0.65 ¡ 0.6 ¡ 1 ¡ 2 ¡ 3 ¡ 4 ¡ 5 ¡ 6 ¡ 7 ¡ 8 ¡ 9 ¡ 10 ¡ 11 ¡ 12 ¡ 13 ¡ 14 ¡ 15 ¡ 16 ¡ 17 ¡

  22. Incremental ¡fusion ¡ 0.16 ¡ 0.14 ¡ 0.12 ¡ Min_NDCR ¡ 0.1 ¡ Min ¡NDCR ¡ 0.08 ¡ 0.0532 ¡ 0.06 ¡ 0.0333 ¡ 0.04 ¡ 0.02 ¡ 0 ¡ 1 ¡ 2 ¡ 3 ¡ 4 ¡ 5 ¡ 6 ¡ 7 ¡ 8 ¡ 9 ¡ 10 ¡ 11 ¡ 12 ¡ 13 ¡ 14 ¡ 15 ¡ 16 ¡ 17 ¡ Number ¡of ¡systems ¡ • We ¡incrementally ¡added ¡systems ¡and ¡computed ¡the ¡fusion ¡ ¡ • Systems ¡5 ¡and ¡15 ¡are ¡the ¡only ¡ones ¡making ¡the ¡fusion ¡worse ¡ • Final ¡Min_NDCR=0.0333 ¡

  23. Fusion ¡of ¡all ¡minus ¡1 ¡ 0.05 ¡ 0.045 ¡ 0.04 ¡ Min_NDCR ¡ 0.035 ¡ 0.03 ¡ Baseline ¡(fusion ¡of ¡all) ¡ 0.025 ¡ 0.02 ¡ 0 ¡ 1 ¡ 2 ¡ 3 ¡ 4 ¡ 5 ¡ 6 ¡ 7 ¡ 8 ¡ 9 ¡ 10 ¡ 11 ¡ 12 ¡ 13 ¡ 14 ¡ 15 ¡ 16 ¡ 17 ¡ We ¡obtain ¡an ¡order ¡from ¡worse ¡to ¡best ¡in ¡the ¡ fusion ¡(worse ¡in ¡here ¡is ¡system ¡15) ¡

  24. Incremental ¡eliminaBon ¡ 0.45 ¡ 0.4 ¡ 0.35 ¡ Min_NDCR ¡ 0.3 ¡ 0.25 ¡ 0.2 ¡ 0.15 ¡ 0.1 ¡ 0.0333 ¡ 0.0685 ¡ 0.0195 ¡ 0.05 ¡ 0 ¡ 0 ¡ 15 ¡ 16 ¡ 11 ¡ 7 ¡ 6 ¡ 14 ¡ 3 ¡ 4 ¡ 9 ¡ 10 ¡ 2 ¡ 17 ¡ 13 ¡ 5 ¡ 1 ¡ 8 ¡ • With ¡only ¡5 ¡systems ¡we ¡achieve ¡prejy ¡decent ¡results ¡ • The ¡best ¡result ¡is ¡0.0195, ¡although ¡this ¡is ¡“cheaBng” ¡

  25. Conclusions ¡ • The ¡fusion ¡algorithm ¡can ¡extract ¡knowledge ¡ and ¡make ¡results ¡bejer ¡ – Even ¡if ¡fusing ¡systems ¡which ¡have ¡weaker ¡NDCR ¡ results, ¡the ¡fusion ¡results ¡in ¡good ¡scores. ¡ • FUTURE ¡WORK: ¡automaBcally ¡idenBfy ¡which ¡ modaliBes ¡bring ¡novelty. ¡

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend


More recommend