BirdCLEF2021まとめ

以下から各リンクに飛べます

English version of this presentation
start (kaggle)
start (twitter)
LAIME
Starter and some thoughts by @hidehisaarai1213
BirdCLEF2021: Exploring the data
BirdCLEF2021: Processing audio data
top solutions and approaches
[1st Place] Quick Solution
[1st Place] Detailed Solution
kami (twitter)
kami (kaggle)
Kaggle の鳥コンペで1位を取った話：BirdCLEF 2021 優勝解法
2nd place solution
4th place solution
Introduction to Sound Event Detection by @hidehisaarai1213
5th place solution
8th place writeup
9th Place solution
My journey (11th solution)
(2020年の鳥コンペ) 18th place solution: efficientnet b3
(Rainforest) 11th place, The 0.931 Magic Explained: Image Classification

Transcript

1. BirdCLEF2021·ͱΊ
   ίϯϖ֓ཁͱ্Ґऀղ๏
   
   
   English version also available
   (ΞΠίϯ୳͠த...)
   start
   

   (@startjapan)
   (Speaker Deckͷ֓ཁཝ͔Β֤ϦϯΫʹඈ΂·͢)
   

   View Slide

2. ࣗݾ঺հ
   

   View Slide

3. ޿ౡͷҩֶੜɽ
   
   
   

   ࠃࢼͷษڧͷ๣Βҩྍϕϯνϟʔ (ג)MNES ʹͯ
   ΠϯλʔϯΛ͓ͯ͠Γɼಉϕϯνϟʔ͕ޙԉ͢Δ
   LAIMEͱ͍͏ֶੜ޲͚ػցֶशษڧαʔΫϧͰ
   

   ݚᮎΛੵΜͰ͍·͢ɽ
   
   
   

   ຊίϯϖͰ༏উ͠ɼMasterͱͳΓ·ͨ͠ɽ
   

   (ଟ෼ʹνʔϜϝΠτͷ͓͔͛Ͱ͕͢...)
   
   
   

   ͋ͱɼྑ͍ΞΠίϯΛ୳͍ͯ͠·͢ɽ
   kaggleɿ@startjapan
   

   twitter ɿ@startjapanml
   ࣗݾ঺հ
   

   View Slide

4. ίϯϖ֓ཁ
   

   View Slide

5. ίϯϖ֓ཁ
   • 5ඵࠁΈͷԻ੠ηάϝϯτ͔Β໐͍͍ͯΔௗछΛಛఆ͢Δίϯϖ
   

   (2020೥ʹ΋ಉ͡ओ࠵ऀ͕ྨࣅίϯϖΛ։࠵͍ͯ͠Δ → 2020೥ͷௗίϯϖͱݺͼ·͢)
   
   
   • trainσʔλ͸xeno-cantoͱ͍͏ௗͷ໐͖੠ڞ༗αΠτ͔Βऔಘ͞ΕͨԻ੠
   

   (train_short_audio)
   
   
   • testσʔλ͸10෼×80݅ͷԻ੠ϑΝΠϧɼ͜ΕΛ5ඵ͝ͱʹ۠੾Γ༧ଌ͢Δ
   

   (test_soundscapes)
   
   
   • ্هͱ͸ผʹvalidation༻ͷԻ੠(10෼×20݅)΋༩͑ΒΕͨ
   

   (train_soundscapes)
   

   View Slide

6. [test_soundscapes]
   
   
   testσʔλɽ10෼×80͕݅ͩఏग़͠ͳ͍ͱΞΫηεͰ͖ͳ͍ɽ
   

   τʔλϧ4ͭͷ৔ॴͰ࿥Ի͞Ε͍ͯΔɽ
   
   
   [train_short_audio]
   
   
   ֶशσʔλɽௗछ͝ͱʹԻ੠͕·ͱΊΒΕ͍ͯΔɽ
   

   ߹ܭͰ62874݅ͷԻ੠σʔλɽ
   
   
   [train_soundscapes]
   
   
   test_soundscapesʹ͍ۙԻڹυϝΠϯΛ࣋ͭɽ
   

   10෼×20݅͋Γɼtest_soundscapesΛ࿥Իͨ͠4ͭͷ৔ॴͷ
   

   ͏ͪ2ͭͷ৔ॴͰ࿥ΒΕͨԻ੠
   
   
   [train_metadata.csv]
   
   
   train_short_audioʹର͢Δmetadataɽshape͸(62784, 14)
   
   
   [train_soundscape_labels.csv / test.csv]
   
   
   10෼ͷϑΝΠϧΛ5ඵηάϝϯτʹ෼͚ͨࡍͷࠎ૊ΈΛఏڙɽ
   

   train_soundscape_labels.csv͸train_soundscapesʹɼ
   

   test.csv͸test_soundscapesʹରԠ͠ɼ
   

   લऀʹͷΈਖ਼ղϥϕϧ͕෇͍͍ͯΔɽ
   
   
   

   View Slide

7. ఏग़ܗࣜ & ධՁࢦඪ
   • 1ͭͷηάϝϯτʹରͯ͠ෳ਺ͷௗछΛ༧ଌͱͯ͠ఏग़Մೳ
   
   
   • ௗ͕໐͍͍ͯͳ͍ηάϝϯτʹରͯ͠͸"nocall"ͱ͍͏จࣈྻΛఏग़
   
   
   • ධՁࢦඪ͸ߦ͝ͱͷmicro-F1είΞͷฏۉ
   

   View Slide

8. 2020೥ͷௗίϯϖͱͷࠩҟ
   

   View Slide

9. 2020೥ͷௗίϯϖͱͷࠩҟ
   • train_soundscapesͷଘࡏ
   

   ʔ train_short_audioͱtestσʔλͷؒʹ͸ԻڹυϝΠϯͷ͕ࠩେ͖͍
   

   ʔ ࠓճͷίϯϖͰ͸ΑΓtestσʔλʹ͍ۙԻڹυϝΠϯΛ࣋ͭtrain_soundscapes͕༩͑ΒΕͨ
   

   ʔ validation༻్Ͱ༻͍ΒΕΔ͜ͱ͕ଟ͔͕ͬͨɼதʹ͸޻෉ֶͯ͠शʹ༻͍Δਓ΋͍ͨ
   
   
   • testσʔλͷҐஔ৘ใʹΞΫηεͰ͖ͨ
   

   ʔ testσʔλͷ֤ϑΝΠϧ໊ʹ৔ॴͷ৘ใ͕ೖ͍ͬͯΔ͜ͱ͸อূ͞Ε͍ͯͨ (೔෇΋)
   

   ʔ ैͬͯɼ͜ΕΒͷ৘ใ΋ԿΒ͔ͷܗͰύΠϓϥΠϯʹ૊ΈࠐΉඞཁ͕͋ͬͨ
   (ࢀߟɿStarter and some thoughts by @hidehisaarai1213)
   

   View Slide

10. EDA (train_short_audioฤ)
    

    View Slide

11. 1ϑΝΠϧ͋ͨΓͷԻ੠ͷ௕͞ (train_short_audio)
    ※ train_short_audioͷ͏ͪ1000݅ ( / 62874݅) ͷԻ੠ϑΝΠϧΛϥϯμϜαϯϓϦϯά
    
    
    ※ ԣ࣠ : 1ϑΝΠϧ͋ͨΓͷԻ੠ͷ௕͞ [ඵ]
    
    
    ※ ॎ࣠ : ౓਺ (߹ܭ1000݅)
    

    View Slide

12. 1छͷௗʹରͯ͠Կ݅ͷԻ੠ϑΝΠϧ͕͋Δʁ (train_short_audio)
    ※ ԣ࣠ : ֤ௗʹ͓͚ΔԻ੠ϑΝΠϧ਺ (train_short_audio಺)
    
    
    ※ ॎ࣠ : ౓਺ (߹ܭ397छ)
    

    View Slide

13. secondary labelsʹ͸ܽଛ͕͋Δͱ໌ه͞Ε͍ͯΔ (train_short_audio)
    (BirdCLEF2021: Exploring the dataΑΓҾ༻)
    

    View Slide

14. EDA (soundscapesฤ)
    

    View Slide

15. શߦnocallఏग़ͰPublicLBͷnocall཰͸෼͔Δ
    BirdCLEF2021
    (ࢀߟɿ2020೥ͷௗίϯϖ)
    Private
    Private Public
    Public
    

    View Slide

16. ҰํͰtrain_soundscapesͰ͸΍΍ߴ͍nocall཰
    

    View Slide

17. train_soundscapesʹ͓͚Δ໨తม਺ͷ෼෍ (nocallࠐΈ)
    • ѹ౗తʹnocall͕ଟ͍
    
    
    • 2छҎ্໐͍͍ͯΔ5ඵηάϝϯτ΋͋Δ
    

    View Slide

18. • Α͘؍ଌ͞ΕΔௗछͷ૊Έ߹Θͤ΋͋Δ
    train_soundscapesʹ͓͚Δ໨తม਺ͷ෼෍ (nocall࡟আ൛)
    

    View Slide

19. train_soundscapesʹ͓͍ͯ5ඵηάϝϯτ಺Ͱಉ࣌ʹ໐͍͍ͯΔௗͷ਺
    

    View Slide

20. Ի੠ೝࣝλεΫͷϕʔγοΫͳղ๏
    

    View Slide

21. Ի੠ೝࣝλεΫͷϕʔγοΫͳղ๏
    Ի੠σʔλ͸ԣ͕࣠࣌ؒɼॎ͕࣠प೾਺ɼ
    

    ֤ϐΫηϧ͕৴߸੒෼ͷڧ౓Λࣔ͢ը૾
    

    (εϖΫτϩάϥϜ) ʹม׵ՄೳͰ͋Γɼ
    

    ͜Εʹରͯ͠CNNͳͲΛదԠ͢Δͱ
    

    ैདྷ௨Γͷը૾ॲཧͱͯ͠ѻ͑Δɽ
    ※ ຊίϯϖͰ͸ॎ࣠(प೾਺)ʹϝϧई౓Λ࢖༻ͨ͠ϝϧεϖΫτϩάϥϜ͕Α͘࢖ΘΕͨ
    

    ※ ϝϧई౓ͱ͸ɿԻͷप೾਺ʹؔͯ͠ɼ͜ͷई౓্Ͱͷ͕ࠩಉ͡Ͱ͋Ε͹ਓ͕ؒࣖͰײ͡ΔԻͷߴ͞ͷࠩ΋ಉ͡ʹͳΔ
    CNN
    (ը૾͸BirdCLEF2021: Processing audio dataΑΓҾ༻)
    

    View Slide

22. ຊίϯϖಛ༗ͷΫη
    

    View Slide

23. ຊίϯϖಛ༗ͷΫη
    • train_short_audioʹରͯ͠weak label͔͠ৼΒΕ͍ͯͳ͍ (weak label໰୊)
    

    ʔ ਺ेඵͷԻ੠σʔλશମʹରͯ͠ϥϕϧ͕෇༩͞Ε͍ͯΔ
    

    ʔ 5ඵ۠੾ΓͷηάϝϯτϨϕϧͰͲͷௗ͕໐͍͍ͯΔ͔͕෼͔Βͳ͍
    
    
    • train_short_audioͷҰ෦Ͱϥϕϧͷܽଛ͕͋Δ (noisy label໰୊)
    

    ʔ ಛʹsecondary_labels(※)ʹ͸ܽଛ͕͋Δͱ໌ه͞Ε͍ͯΔ
    
    
    • ࿥Ի೔΍৔ॴͷ৘ใͳͲͷmetadata΋ԿΒ͔ͷܗͰ৫ΓࠐΉඞཁ͕͋Δ (metadataͷ৫ࠐ)
    
    
    • ༧ଌର৅ͷલޙͷηάϝϯτͰௗ͕໐͍͍ͯΔ͔ͱ͍͏৘ใ΋ҙຯΛ࣋ͭՄೳੑ͕͋Δ
    

    (ηάϝϯτલޙ৘ใͷ৫ࠐ)
    
    
    • train_soundscapesͱtest_soundscapesͰnocall཰͕େ͖͘ҟͳΔ (CVઓཱུ֬ͷࠔ೉)
    ※ train_short_audioͷϥϕϧʹ͸primary_labelͱsedondary_labelsͷ2छྨ͕͋Δ
    

    View Slide

24. ্Ґऀͷղ๏
    top solutions and approaches
    
    
    ্هͷdiscussionʹ্Ґऀղ๏΁ͷϦϯΫ͕·ͱ·͍ͬͯ·͢
    

    View Slide

25. 1st place (ours!)
    [1st Place] Quick Solution
    
    
    [1st Place] Detailed Solution
    

    View Slide

26. tl;dr
    1st stage : ֎෦σʔλ(freefield1010)Λ࢖ͬͯbinary nocall detector ࡞੒ (1 : Կ͔ௗ໐͍ͯΔ / 0 : nocall)
    

    2nd stage : nocall detectorΛ࢖ͬͯtrain_short_audio͔Βnocall෦෼ͷweightΛݮΒ্ͨ͠Ͱ397࣍ݩϚϧνϥϕϧ෼ྨثΛ࡞੒
    

    3rd stage : nocall detectorͷ݁Ռɼmetadataɼ2nd stageͷ݁ՌͳͲ͔Βࣗલtable competitionΛ࡞੒
    
    
    ࠷ऴతʹࣗલtable competitionʹ͢Δ͜ͱͰ
    

    weak label໰୊ɼnoisy label໰୊ɼmetadataͷ৫ࠐɼηάϝϯτલޙ৘ใͷ৫ࠐͳͲΛ·Δͬͱղܾʂʂ
    ※ Inference Part ͷΈͷུ֓Ͱ͋Γɼ1st stage෦෼͸লུ͍ͯ͠·͢
    

    View Slide

27. ͳͥtableԽͰweak label໰୊ & noisy label໰୊͕ղܾ͞ΕΔʁ
    • 3rd stageͷtargetม਺ (0 : ͸ͣΕߦ / 1 : ͋ͨΓߦ) ͸ҎԼͷྲྀΕͰܾఆ͞ΕΔ
    • ਺ेඵͷԻ੠σʔλʹରͯ͠෇༩͞Εͨprimary & secondary labelsʹରͯ͠ηάϝϯτ୯ҐͰ༧ଌ஋Λग़ͤΔ
    

    nocall detectorͱϚϧνϥϕϧ෼ྨثͷग़ྗΛ૊Έ߹ΘͤΔ͜ͱͰweak label໰୊Λղܾ
    
    
    • Ծʹsecondary labelsʹܽଛ͕͋Δͱϥϕϧ0͕෇༩͞ΕΔ͕ϥϕϧ0ͷαϯϓϧ਺͸ൺֱతଟ͘
    

    noise͸͍͍ײ͡ʹຒ΋ΕΔ (noisy label໰୊ͷ؇࿨)
    

    View Slide

28. more details...
    • νʔϜϝΠτͷkami͞Μ (twitter : @634kami / kaggle : @kami634) ͕ҎԼʹ೔ຊޠͰղ๏Λ·ͱΊͯ͘Ε·ͨ͠
    

    Kaggle ͷௗίϯϖͰ1ҐΛऔͬͨ࿩ɿBirdCLEF 2021 ༏উղ๏
    

    View Slide

29. 2nd place
    2nd place solution
    

    View Slide

30. (2nd place solutionΑΓҾ༻)
    2nd place
    

    View Slide

31. 2nd place
    • train_short_audio͔Β30ඵ୯ҐͰநग़ͨ͠ͷͪɼ5ඵ͝ͱʹ۠੾Γɼmixup͢Δ
    

    (weak label໰୊΁ͷରԠ)
    
    
    • train_soundscapesͷ͏ͪ10෼ؒશ͘ௗ͕໐͔ͳ͍Ի੠ϑΝΠϧ3ͭͷআ֎ & ϒʔτετϥοϓαϯϓϦϯά
    

    (ϩόετͳCVઓུ)
    
    
    • label smoothing & metadataதͷratingྻΛ༻͍ͯॏΈ෇͚ (noisy label໰୊΁ͷରԠ)
    
    
    • ᮢ஋બ୒ͷࡍͷtips
    

    ʔ LBͰ͸CVΑΓnocall཰͕௿͍ͷͰᮢ஋ΛԼ͛ͯௗΛଟ͘༧ଌ
    

    ʔ ϞσϧؒͰ֬཰஋ͷ෼෍͕ҟͳΔͨΊ୯Ұͷ֬཰஋Λᮢ஋ͱ͢Δͷ͸φϯηϯε
    

    ΑͬͯɼύʔηϯλΠϧϕʔεͷᮢ஋Λ࢖༻
    
    
    • ͦͷଞ (ޙॲཧ)
    

    ʔ ௗ͝ͱͷฏۉ༧ଌ֬཰͔Βݸʑͷ֬཰஋Λमਖ਼
    

    ʔ લޙηάϝϯτ৘ใΛ࢖༻
    

    ʔ nocall detectorͷ݁ՌΛՃຯ
    

    ʔ ࣌ͱ৔ॴͷ৘ใ͔Β͋Γ͑ͳ͍ௗछΛ༧ଌΛ͍ͯ͠Δ৔߹͸࡟আ (metadataͷ৫ࠐ)
    

    View Slide

32. 4th place
    4th place solution
    

    View Slide

33. 4th place
    • SEDϞσϧΛ࢖༻ɼೖྗ͸10-30ඵ (weak label໰୊΁ͷରԠ)
    

    (ࢀߟɿIntroduction to Sound Event Detection by @hidehisaarai1213)
    
    
    • ͜ͷํ΋mixupΛ࢖༻
    
    
    • psudo labelingΛ࣮ߦ (noisy label໰୊΁ͷରԠ)
    
    
    • ༧ଌର৅ͷ5ඵηάϝϯτͱͦΕΛத৺ͱ͢Δ30ඵηάϝϯτͷͦΕͧΕʹର͢ΔSEDͷग़ྗ
    Λ૊Έ߹Θͤͯ࠷ऴग़ྗͱͨ͠ (ηάϝϯτલޙ৘ใͷ৫ࠐ)
    

    ʔ 5ඵηάϝϯτʹରͯ͠͸খ͞ͳᮢ஋ɼ30ඵηάϝϯτʹରͯ͠͸େ͖ͳᮢ஋Λ࢖༻
    
    
    • 2Ґͷղ๏ͱಉ༷ʹɼ࣌ͱ৔ॴͷ৘ใ͔Β؍ଌ͞ΕΔՄೳੑ͕௿͍ͱ൑அͨ͠ௗछ͸࡟আ
    

    (metadataͷ৫ࠐ)
    

    View Slide

34. 5th place
    5th place solution
    

    View Slide

35. 5th place
    • 2020೥ͷௗίϯϖͰ2ҐͩͬͨํͰ͋Γɼࠓճ΋ͦΕΛϕʔεͱ͍ͯͨ͠
    
    
    • લճ͔Βͷվળ఺ɿSEDʹมߋͰ +1% (※1) / ᮢ஋ௐ੔ʹΑΓ +1% / Ξϯαϯϒϧํ๏վྑͰ +1%
    

    (+ ஍Ҭ৘ใΛ΋ͱʹ༧ଌϥϕϧͷߜΓࠐΈ΋ͨ͠Έ͍ͨ(※2) )
    
    
    • augmentation͕ಛ௃తɿը૾Λ0.5-3৐ / nഒ଎ / Ӎ΍ձ࿩ͳͲͷԻΛ௥Ճ / ϊΠζ௥Ճ / 0.5ͷ֬཰Ͱप೾਺ௐ੔
    

    (1-4Ґ͸mixup΍ϊΠζ௥Ճʹཹ·Δҹ৅)
    
    
    • primary label͸ϥϕϧ1, secondary labels͸ϥϕϧ0.3Λ෇༩
    
    
    • 1ͭͷηάϝϯτͰ؍ଌ͞Εͨௗ͸10෼ͷԻ੠ϑΝΠϧશମͰर্͍͛΍͘͢ͳΔΑ͏ௐ੔ (※3)
    ※1 : weak label໰୊΁ͷରԠ
    
    
    ※2 : metadataͷ৫ࠐ
    
    
    ※3 : ηάϝϯτલޙ৘ใͷ৫ࠐ
    

    View Slide

36. 8th place
    8th place writeup
    

    View Slide

37. 8th place
    • 2020೥ͷௗίϯϖͰ6Ґͩͬͨํɼࠓճ΋SEDΛ࢖༻ (weak label໰୊΁ͷରԠ)
    
    
    • ֶश࣌͸5ඵ or 20ඵηάϝϯτɼਪ࿦࣌͸40ඵηάϝϯτΛ࢖༻ɼ௕͍΄͏͕Α͔ͬͨ
    

    ·ͨɼਪ࿦͸0-40ඵͰߦͬͨ࣍ʹ20-60ඵͱ͍͏෩ʹoverlapΛ΋ͨͤͨ (ηάϝϯτલޙ৘ใͷ৫ࠐ)
    
    
    • augmentationɿΨ΢γΞϯϊΠζɼϐϯΫϊΠζɼϘϦϡʔϜௐ੔ɼϐονγϑτ
    

    (mixup΋্ख͘ߦ͕ͬͨܭࢉࢿݯͷ໰୊Ͱ࠷ऴఏग़ʹ͸૊ΈࠐΊͳ͔ͬͨͦ͏)
    
    
    • ଛࣦؔ਺͕ಛ௃త (BCEFocal2WayLoss)
    
    
    • primary labelͱsecondary labels͸ಉ͡Α͏ʹѻͬͨ
    
    
    • psudo labelingΛ࣮ߦ (noisy label໰୊΁ͷରԠ)
    
    
    • ᮢ஋͸call thresholdͱnocall thresholdͷ2͕ͭଘࡏ͠ɼcall thresholdΛ௒͑ͨௗछ͸ཅੑͱ͢ΔҰํͰ
    

    શͯͷௗछʹ͓͍ͯnocall thresholdΛ௒͑ͳ͔ͬͨηάϝϯτʹ͸nocall΋෇༩ (ௗϥϕϧͱnocall͕ڞଘ͠͏Δ)
    
    
    • ஍Ҭ৘ใ͔Βଘࡏ͢Δ͸͕ͣͳ͍ௗछ͸༧ଌ͍ͯͯ͠΋আ֎ (metadataͷ৫ࠐ)
    
    
    • ௗ͕໐͍͍ͯΔߦͱnocallߦʹ෼͚ͯF1είΞΛࢉग़͠0.54 * nocall_f1 + 0.46 * call_f1ͰCVΛಋग़ (ϩόετͳCVઓུ)
    

    View Slide

38. 9th place
    9th Place solution
    

    View Slide

39. 9th place
    • ֶश࣌ͷೖྗ͸5-7ඵηάϝϯτ
    
    
    • secondary labelsͷॏΈ͸খͨ͘͞͠
    
    
    • mixup࢖༻
    
    
    • ՄೳͳݶΓͷଟ༷ੑΛ΋ͨͤͨ
    

    ʔ ࣌ؒ෼ղೳͷҟͳΔmel-spectrogramɼhop_length͸200ͱ320
    

    ʔ ༷ʑͳbackbone
    

    ʔ augmentationɿwhite noise, pink noise, band noise, nocall clipsɼmel-spectrogramը૾ͷྦྷ৐
    
    
    • ޙॲཧ
    

    ʔ 10෼ͷԻ੠σʔλશମʹ͓͚Δ֤ௗͷ໐͘࠷େ֬཰ or ฏۉ֬཰Ͱޙॲཧ (ηάϝϯτલޙ৘ใͷ৫ࠐ)
    

    ʔ ஍Ҭ৘ใ͔ΒͲΕ͚֤ͩௗ͕໐͘Մೳੑ͕͋Δ͔ධՁͯͦ͠ͷ݁ՌͰޙॲཧ (metadataͷ৫ࠐ)
    

    ʔ 1೔ͷؒͰ֤ௗ͕໐͘࠷େ֬཰Λ࢖ͬͯޙॲཧ (metadataͷ৫ࠐ)
    
    
    • Squeeze width of test soundscapes by 2-5% (mostly to reverse far field effects)
    
    
    • ͜ͷํ΋ᮢ஋Λ2ͭ(call, nocall)ઃఆ͠ɼௗϥϕϧͱnocallͷڞଘΛೝΊͨ
    

    View Slide

40. 11th place
    My journey (11th solution)
    

    View Slide

41. 11th place
    • Public LBͰ௕͍͜ͱटҐΛಠ઎͞Ε͍ͯͨCPMP͞Μ
    
    
    • 2020೥ͷௗίϯϖͰ18ҐɼRainforestίϯϖͰ11ҐΛͱΒΕͨํͰ͋Γ྆ऀͷղ๏Λmixͨ͠΋ͷΛϕʔεͱͨͦ͠͏
    

    ʔ 2020೥ͷௗίϯϖͷղ๏ : 18th place solution: efficientnet b3
    

    ʔ Rainforestίϯϖͷղ๏ : 11th place, The 0.931 Magic Explained: Image Classification
    
    
    • 8Ґͷղ๏ͱಉ͘͡0.54 * nocall_f1 + 0.46 * call_f1ʹͯCVΛࢉग़ (ϩόετͳCVઓུ)
    
    
    • ΞϯαϯϒϧͰ࣮֬ʹείΞ্͕ঢ͢Δͱա৴͓ͯ͠Γίϯϖऴྃ਺೔લ·ͰΞϯαϯϒϧverΛఏग़͠ͳ͔ͬͨ͜ͱ
    Λޙչͳ͍ͬͯ͞Δ (࣮ࡍʹ͸Ξϯαϯϒϧ͕ޮ͔ͳ͔ͬͨͦ͏)
    

    View Slide

42. ͋Γ͕ͱ͏͍͟͝·ͨ͠ʂ
    

    View Slide