Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
iclr2020deepsemi-supervisedanomalydetectionyama...
Search
Yamato.OKAMOTO
June 14, 2020
Technology
0
84
iclr2020deepsemi-supervisedanomalydetectionyamatookamoto-200531022507.pdf
Yamato.OKAMOTO
June 14, 2020
Tweet
Share
More Decks by Yamato.OKAMOTO
See All by Yamato.OKAMOTO
いまAI組織が求める企画開発エンジニアとは?
roadroller
2
1.3k
Slide ICCV2023 Constructing Image Text Pair Dataset from Books
roadroller
0
46
第11回 全日本コンピュータビジョン勉強会 CVPR2022 "A Self-Supervised Descriptor for Image Copy Detection"
roadroller
0
520
第9回 全日本コンピュータビジョン勉強会 発表資料
roadroller
0
560
第七回全日本コンピュータビジョン勉強会 A Multiplexed Network for End-to-End, Multilingual OCR
roadroller
1
890
部下のマネジメントはAI開発に学べ
roadroller
0
95
Domain Generalization via Model-Agnostic Learning of Semantic Features NeurIPS’19 読み会 in 京都
roadroller
0
190
ICML’2019 読み会in京都 Federated Learningの研究動向
roadroller
0
50
CVPR2019@Long Beach 参加速報(本会議)
roadroller
0
77
Other Decks in Technology
See All in Technology
適材適所の技術選定 〜GraphQL・REST API・tRPC〜 / Optimal Technology Selection
kakehashi
1
150
【Pycon mini 東海 2024】Google Colaboratoryで試すVLM
kazuhitotakahashi
2
480
なぜ今 AI Agent なのか _近藤憲児
kenjikondobai
4
1.3k
Amazon CloudWatch Network Monitor のススメ
yuki_ink
1
200
AGIについてChatGPTに聞いてみた
blueb
0
130
Why App Signing Matters for Your Android Apps - Android Bangkok Conference 2024
akexorcist
0
120
The Rise of LLMOps
asei
5
1.1k
リンクアンドモチベーション ソフトウェアエンジニア向け紹介資料 / Introduction to Link and Motivation for Software Engineers
lmi
4
300k
【令和最新版】AWS Direct Connectと愉快なGWたちのおさらい
minorun365
PRO
5
740
ドメインの本質を掴む / Get the essence of the domain
sinsoku
2
150
透過型SMTPプロキシによる送信メールの可観測性向上: Update Edition / Improved observability of outgoing emails with transparent smtp proxy: Update edition
linyows
2
210
安心してください、日本語使えますよ―Ubuntu日本語Remix提供休止に寄せて― 2024-11-17
nobutomurata
0
980
Featured
See All Featured
Responsive Adventures: Dirty Tricks From The Dark Corners of Front-End
smashingmag
250
21k
Save Time (by Creating Custom Rails Generators)
garrettdimon
PRO
27
840
The Straight Up "How To Draw Better" Workshop
denniskardys
232
140k
Typedesign – Prime Four
hannesfritz
40
2.4k
Designing Dashboards & Data Visualisations in Web Apps
destraynor
229
52k
10 Git Anti Patterns You Should be Aware of
lemiorhan
654
59k
How STYLIGHT went responsive
nonsquared
95
5.2k
Art, The Web, and Tiny UX
lynnandtonic
297
20k
Understanding Cognitive Biases in Performance Measurement
bluesmoon
26
1.4k
Dealing with People You Can't Stand - Big Design 2015
cassininazir
364
24k
Build your cross-platform service in a week with App Engine
jlugia
229
18k
Become a Pro
speakerdeck
PRO
25
5k
Transcript
2020/6/14 Yamato OKAMOTO ICLRΦϯϥΠϯಡΈձ Deep Semi-supervised Anomaly Detection
ࣗݾհʢ͘!!ʣ ɹԬຊେʢ͓͔ͱ·ͱʣ • ژେֶඒೱݚڀࣨͰύλʔϯೝࣝΛݚڀͯ͠म࢜՝ఔमྃ • ΦϜϩϯͰ৽نࣄۀΛܦݧޙɺ͍·ࣾձγεςϜࣄۀ෦ͷݚڀॴϦʔμʔ • ເژΛϙετɾγϦίϯόϨʔʹ͢Δ͜ͱɺؔͷίϛϡχςΟΛڧԽ͍ͨ͠ ɹ㱺 ژͷมਓύϫʔΛੈքʹΒ͠Ί͍ͨ
Twitter : RoadRoller_DESU ҆৺҆શͳࣾձͷ࣮ݱʹ͚ͯɺ ࠷ۙ Anomaly Detection ʹڵຯΞϦ
Anomaly Detection ͋Δ͋Δ ఆٛࠔ • ҟৗʹ༷ʑͳόϦΤʔγϣϯ͕͋Δ • ҟৗݕग़͍͚ͨ͠ͲʮWhat is ҟৗʁʯ͕ఆٛͰ͖ͳ͍
ֶशσʔλ͕ೖखࠔ • ҟৗ໓ଟʹൃੜ͠ͳ͍ʢ※ සൟʹൃੜ͢ΔΠϕϯτҟৗ͡Όͳ༷ͯ͘ʣ • ѹతʹҟৗσʔλ͕ෆͯ͠ػցֶश͕ࠔ ैདྷख๏ɿਖ਼ৗΛఆٛ͢Δ • ʮWhat is ҟৗʁʯͷఆٛΛఘΊΔɺҟৗσʔλͷֶशఘΊΔ • ͦͷΘΓʮWhat is ਖ਼ৗʁʯͷఆٛΛֶशͯ͠ɺʮNot ਖ਼ৗʯΛҟৗͱఆ͢Δ
Anomaly Detection ͷैདྷݚڀ Deep One-Class Classification (ICML’18) • ਖ਼ৗσʔλͷΈΛ༻͍ͯɺClassifierͳΓAutoEncoderͳΓΛैདྷ௨Γʹֶश •
͜ͷͱ͖ɺಛྔ͕࣍ݩ෦ۭؒʹऩଋ͢ΔΑ͏LOSSΛՃ͑Δ • ਖ਼ৗσʔλͳΒٿʹ͢ΔͣͳͷͰɺٿ͔Β֎ΕͨσʔλΛҟৗͱఆ͢Δ ୈҰ߲ʹΑͬͯٿʹ͕ԡ͠ࠐ·ΕΔ cɿ ٿͷத৺ʢͨͩ͠≠0ʣ nɿֶश͢Δਖ਼ৗσʔλͷ
Anomaly Detection ͷධՁ؍ ͲΕ͚ͩਖ਼֬ʹҟৗΛݕͰ͖͔ͨʁ • ਖ਼ৗσʔλΛਖ਼ৗͱఆͯ͠ɺҟৗσʔλΛҟৗͱఆ͢Δਫ਼ ԼྲྀλεΫΛअຐ͠ͳ͍͔ʁ • ԼྲྀλεΫ͕͋Δ߹ɺҟৗݕػೳͷՃʹΑͬͯѱӨڹ͕ͳ͍͔Ͳ͏͔ •
ྫ͑ɺ10ΫϥεͷࣈࣝผثʹɺਤܗͳͲࣈҎ֎͕ೖྗ͞Εͨͱ͖ҟৗͱఆ͢Δػ ೳΛ͚Ճ͍͑ͨͤͰɺैདྷͷ10Ϋϥεࣝผੑೳ͕Լ͢ΔͱࠔΔ ad-hoc͔post-hoc͔ʁ • ҟৗݕ͢ΔͨΊʹϞσϧߏֶशํ๏·Ͱม͑Δඞཁ͕͋Δ͔ʁ • ·ͨɺLOSSΛޙ͔Β͚͚̍ͭͩͯ͠Ճֶश͢Δ͚ͩͰOK͔ʁ • ͲͪΒ͕ྑ͍ѱ͍ͳͲҰ֓ʹݴ͑ͳ͍͕ɺpost-hocͷํ͕ѻ͍͍͢ɻ
հจͷ֓ཁ ʮSemi-supervisedʹֶश͠Α͏ʂʯ Anomaly Detection ͷݚڀUnsupervised͕ओྲྀͷΑ͏ͩ Ͱɺֶश༻ͷҟৗσʔλ͕ೖखࠔͩͱͯ͠ɺ ӡ༻Λଓ͚ͯͨΒҟৗσʔλʹ͍ͣΕग़ձ͏ͣ ͳΒɺͦΕΒগྔͷҟৗσʔλΛͬͯɺ Semi-supervisedʹֶशͨ͠ํ͕ྑ͍ͷͰʁ ※Semi-supervisedͷAnomaly
Detectionݚڀඇৗʹগͳ͍
ఏҊख๏ ʮLOSSʹ߲Λ̍ͭՃ͠·ͨ͠ʯ Deep One-Class Classification (ICML’18) ͷLOSSʹSemi-supervisedͷ߲Λ̍ͭՃ • ࣮ಉ͡ஶऀͰͨ͠ɻࣗͷݚڀΛࣗͰΞοϓσʔτͨ͠ܗʹͳΔɻ ͠ҟৗσʔλʹग़ձͬͨΒɺ
ٿͷ֎ଆʹߦ͘Α͏ֶश͢Δ mɿsemi-supervisedʹֶश͢Δσʔλ yj ɿਖ਼ৗorҟৗͷϥϕϧ
࣮ݧ݁Ռ ॎ࣠ɿҟৗσʔλͷݕग़ੑೳ ʢHigher is Betterʣ Unsupervised Semi-supervised ԣ࣠ɿSemi-supervisedͰڭࢣ͖ͷҟৗσʔλΛֶशׂͨ͠߹ ఏҊख๏ MNISTɺFashion-MNISTɺCIFAR-10ͷσʔληοτͰධՁ
• ̍Ϋϥεͱਖ਼ৗͱఆٛͯ͠ɺAutoEncoderʴఏҊख๏ͰಛྔදݱΛֶश • Γͷ̕ΫϥεΛೖྗͨ͠ͱ͖ɺҟৗͱఆͰ͖Δ͔Ͳ͏͔ධՁ ੑೳվળΛ֬ೝ
·ͱΊͱߟ ਂֶशʹΑΔ Semi-supervised ͳ Anomaly Detection ख๏ΛఏҊ • ॳΊͯͰͳ͍ͱࢥ͏͕ɺਂֶशʹΑΔAnomaly DetectionͰsemi-supervised͍͠
• ͔ͨ͠ʹࣾձ࣮Λߟ͑Δͱɺ͜ͷઃఆద • ख๏γϯϓϧͰɺpost-hocͳͷͰѻ͍͍͢ • ࠓճԼྲྀλεΫ͕AE͕ͩͬͨɺClassificationͩͱͲ͏ͳΔ͔ʁ • Anomaly DetectionͷධՁσʔληοτͬͯଞʹͳ͍ͷ͔ͳɺɺɺɺ ʢ͍ͭ·ͰMNISTʹΑΔධՁ͕ଓ͘ͷͩΖ͏͔ʣ