DevFest Tokyo 2017 TensorFlow

C-LIS CO., LTD.

C-LIS CO., LTD. ༗ࢁܓೋʢ,FJKJ"3*:"."ʣ $-*4$0 -5% Photo : Koji MORIGUCHI
(AUN CREATIVE FIRM) "OESPJEΞϓϦ։ൃνϣοτσΩϧ ػցֶश͸ͪΐͬͱ΍ͬͨ͜ͱ͋Γ·͢ ΍ͬͯ·ͤΜ

%FW'FTU5PLZP 5FOTPS'MPXͰ  झຯͷը૾ऩूαʔόʔΛ࡞Δ ೥݄߸

લճ·Ͱͷ͓࿩

5FOTPS'MPXൃදʢ೥݄ʣ

ษڧձ΍Ζ͏ͥ

(PPHMF%FWFMPQFS(SPVQ

IUUQTHEHLPCFEPPSLFFQFSKQFWFOUT

Πϯλʔωοτ͔Β ޷Έͷը૾ΛࣗಈͰऩू͍ͨ͠

γεςϜߏ੒ Downloader Face Detection Megane Detection ֬ೝɾमਖ਼ ೝࣝ݁Ռ JSON
ֶशʢ܇࿅ʣ λΠϜϥΠϯ ϝσΟΞ σʔληοτ ֶशʢ܇࿅ʣ TensorFlow

{ "generator": "Region Cropper", "file_name": "haruki_g17.png", "regions": [ { "probability":
1.0, "label": 2, "rect": { "left": 97.0, "top": 251.0, "right": 285.0, "bottom": 383.0 } }, { "probability": 1.0, "label": 2, "rect": { "left": 536.0, "top": 175.0, "right": 730.0, "bottom": 321.0 } } ] } Region Cropper: https://github.com/keiji/region_cropper

؟ڸ່ͬ൑ఆ conv 3x3x3 32 conv 3x3x32 64 max_pooling lrn conv
3x3x64  128 conv 3x3x128 128 max_pooling lrn conv 3x3x128  128 conv 3x3x128 256 fc 384 fc 256 output 2

ͷ൑ఆਫ਼౓ 1 0

ٕज़ܥಉਓࢽΦϯϦʔΠϕϯτ ٕज़ॻయ̍ˍ̎ https://techbookfest.org/tbf01/#A-28 https://techbookfest.org/event/tbf02/circle/5705718560718848

੔ཧ଴ͪͷը૾ ຕ ʢ2017.10.08 ࣌఺ʣ

σʔληοτͷ੔උʢ੔ཧʣ͕  ௚ۙͷ՝୊

Πϥετإݕग़ʢ'BDF%FUFDUJPOʣ

γεςϜߏ੒ Downloader Face Detection Megane Detection ֬ೝɾमਖ਼ ೝࣝ݁Ռ JSON
ֶशʢ܇࿅ʣ λΠϜϥΠϯ ϝσΟΞ σʔληοτ ֶशʢ܇࿅ʣ TensorFlow

3$// 3FHJPOCBTFE$POWPMVUJPOBM/FVSBM/FUXPSLT

4MJEJOH8JOEPX

4FMFDUJWF4FBSDI 4FMFDUJWF4FBSDIGPS0CKFDU3FDPHOJUJPO IUUQXXXIVQQFMFOOMQVCMJDBUJPOTTFMFDUJWF4FBSDI%SBGUQEG 848

%FW'FTU5PLZP 5FOTPS'MPXͰ Πϥετإݕग़ثΛ࡞Δ

ֶशσʔλͷ࡞੒

࠲ඪͷਖ਼ଇԽɹਖ਼نԽ 888x613 left : 536 → 0.6036 top : 175
→ 0.2854 right : 730 → 0.8220 bottom : 321 → 0.5236 0.1092 ← 97 : left 0.4094 ← 251 : top 0.3209 ← 285 : right 0.6247 ← 383 : bottom

ը૾ͷϦαΠζɾάϨʔεέʔϧԽ 256x256 left : 0.6036 top : 0.2854 right :
0.8220 bottom : 0.5236 0.1092 : left 0.4094 : top 0.3209 : right 0.6247 : bottom

Ϩίʔυߏ଄ <JNBHF>ը૾σʔλʢόΠτྻʣ <MBCFMT>
<SFHJPOT>     

XJUIUGQZUIPO@JP5'3FDPSE8SJUFS PVUQVU@QBUI PQUJPOT BTXSJUFS FYBNQMFUGUSBJO&YBNQMF GFBUVSFTUGUSBJO'FBUVSFT GFBUVSF\ GJMF@OBNF@CZUFT@GFBUVSF GJMF@OBNFFODPEF
JNBHF@CZUFT@GFBUVSF JNBHF@EBUB SFHJPO@DPVOU@JOU@GFBUVSF SFHJPO@DPVOU SFHJPOT@GMPBU@GFBUVSF GMBUUFO@SFHJPOT MBCFMT@JOU@MJTU@GFBUVSF MBCFMT ^ XSJUFSXSJUF FYBNQMF4FSJBMJ[F5P4USJOH

Ϟσϧ conv 5x5x1  128 batch_normalization conv 1x1x128  64 conv 3x3x64 
64 batch_normalization conv 1x1x64  128 ReLU conv 1x1x128  128 conv 3x3x128  128 batch_normalization conv 1x1x128  128 ReLU conv 1x1x256  256 conv 3x3x256  256 batch_normalization conv 1x1x256  128 ReLU conv 2x2x128  128 fc 2048 fc 4 256x256x1 Residual block 1 Residual block 2 Residual block 3 × 3 × 3 × 3 avg_pool 2x2 avg_pool 2x2 avg_pool 2x2 ReLU Sigmoid

ೖྗͱग़ྗ 256x256x1 Ϟσϧ Sigmoid

ReLU Sigmoid ׆ੑԽؔ਺

256x256 left : 0.6036 top : 0.2854 right : 0.8220
bottom : 0.5236 0.1092 : left 0.4094 : top 0.3209 : right 0.6247 : bottom ࠲ඪͷ஋Ҭʢʣ

ReLU Sigmoid ׆ੑԽؔ਺

܇࿅ɾֶश ϛχόον ֶश཰ʢ"EBN0QUJNJ[FSʣ

ݕग़݁Ռ

̎ਓҎ্͍Δ৔߹͸ʁ

σʔλ಺༁ 0 (18.57%) 1 (57.64%) 2 (11.93%) إͷ਺ ϑΝΠϧ਺
ׂ߹ 0 3748 18.57% 1 11636 57.64% 2 2408 11.93% 3 844 4.18% 4 596 2.95% 5 325 1.61% 6 214 1.06% 7 108 0.53%

ೖྗͱग़ྗ 256x256x1 Ϟσϧ Sigmoid

إྖҬͷݕग़

إྖҬͷݕग़ dconv 3x3x1 64 dconv 3x3x64 64 dconv 3x3x256 128
dconv 3x3x64 128 dconv 3x3x128 128 conv 3x3x1 64 conv 3x3x64 64 conv 3x3x128 256 conv 3x3x64 128 conv 3x3x128 128

إྖҬͷݕग़

... ৞ΈࠐΈ૚ʢDPOWEʣͷ໾ׂ ಛ௃Ϛοϓ

إྖҬͷݕग़ʢ৞ΈࠐΈΦʔτΤϯίʔμʣ dconv 3x3x3 64 dconv 3x3x64 64 dconv 3x3x256 128
dconv 3x3x64 128 dconv 3x3x128 128 conv 3x3x1 64 conv 3x3x64 64 conv 3x3x128 256 conv 3x3x64 128 conv 3x3x128 128

৞ΈࠐΈ૚ͷΈʹͨ͠৔߹ conv 3x3x1 64 conv 3x3x64 64 conv 3x3x128 256
conv 3x3x64 128 conv 3x3x128 128 dconv 3x3x3 64 dconv 3x3x64 64 dconv 3x3x256 128 dconv 3x3x64 128 dconv 3x3x128 128

৞ΈࠐΈ૚ͷΈʹͨ͠৔߹ conv 3x3x64 stride 1 conv 3x3x64  stride 1 ReLU
ReLU conv 3x3x128  stride 1 conv 3x3x128  stride 1 ReLU conv 3x3x256  stride 1 conv 3x3x256  stride 1 ReLU conv 1x1x1  stride 1 256x256x1 max_pool 2x2 stride 2 max_pool 2x2 stride 2 ReLU ReLU Sigmoid 29x29x1

άϦουʢYʣ

άϦουͷͲ͜ʹإ͕͋Δ͔

࠲ඪͷதԝͷηϧΛબ୒

֤ηϧʹσʔλΛઃఆ confidence : 1.0 center_x : 0.7128 center_y :
0.4045 width : 0.2184 height : 0.2382 1.0 : confidence 0.2150 : center_x 0.5169 : center_y 0.2117 : width 0.2151 : height confidence : 0.0 center_x : 0.0 center_y : 0.0 width : 0.0 height : 0.0

Ϩίʔυߏ଄ <JNBHF>ը૾σʔλʢόΠτྻʣ <CPY@MBCFMT>
<CPY@SFHJPOT>     

Ϟσϧ conv 3x3x64 stride 1 conv 3x3x64  stride 1 ReLU

ग़ྗ conﬁdence center_x center_y width height 29 × 29
× (4 + 1) = 4,205

܇࿅ɾֶश ࠷దԽΞϧΰϦζϜ"EBN0QUJNJ[FS ֶश཰

Yͷݶք إதԝҎ֎ͷηϧͷDPOpEFODF΋ߴ͍ DPOpEFODFͷߴ͍ηϧ͔Β  ਖ਼͍͠࠲ඪ͕ಘΒΕͳ͍

৞ΈࠐΈ૚ͷ໾ׂ conv 3x3x64 stride 1 conv 3x3x64  stride 1 ReLU

... લ૚͔ΒͷೖྗΛεΩϟϯͯ͠ಛ௃ϚοϓΛग़ྗ ಛ௃Ϛοϓ

ಛ௃Ϛοϓ͸ɺ খ͘͞ͳΔ΄ͲେҬͷಛ௃Λද͢

దͨ͠αΠζͷηϧʹׂΓ౰ͯΔ

৞ΈࠐΈΛ܁Γฦ͢աఔͰग़ྗ͢Δ Single Shot MultiBox Detector https://arxiv.org/abs/1512.02325

Ϟσϧ

conv 3x3x64 stride 1 conv 3x3x64  stride 1 ReLU ReLU
conv 3x3x128  stride 1 conv 3x3x128  stride 1 ReLU conv 3x3x256  stride 1 conv 3x3x256  stride 1 ReLU 256x256x1 max_pool 2x2 stride 2 max_pool 2x2 stride 2 ReLU ReLU 29x29x256

conv 3x3x256  stride 2 conv 3x3x256  stride 2 conv 3x3x256 
stride 2 conv 1x1x5  stride 1 Sigmoid conv 1x1x5  stride 1 conv 1x1x5  stride 1 Sigmoid conv 1x1x5  stride 1 Sigmoid conv 1x1x5  stride 1 Sigmoid 1x1x5 15x15x5 8x8x5 2x2x5 4x4x5 29x29x256 conv 1x1x5  stride 1 Sigmoid conv 3x3x256  stride 2 conv 3x3x256  stride 2 Sigmoid 29x29x5

ֶशσʔλ

ֶशσʔλ confidence : 1.0 center_x : 0.7128 center_y :
0.4045 width : 0.2184 height : 0.2382 1.0 : confidence 0.2150 : center_x 0.5169 : center_y 0.2117 : width 0.2151 : height confidence : 0.0 center_x : 0.0 center_y : 0.0 width : 0.0 height : 0.0

ηϧͷׂΓ౰ͯج४ ÷ = Jaccard Overlap ҰఆҎ্ͷ஋ʢॏͳΓʣͷηϧʹׂΓ౰ͯ ↓ + - )
(

EFG@MBOE SFDU XJEUISFDU<*/%&9@3*()5>SFDU<*/%&9@-&'5> IFJHIUSFDU<*/%&9@#0550.>SFDU<*/%&9@501> JGXJEUIPSIFJHIU SFUVSO SFUVSOXJEUI IFJHIU EFGKBDDBSE@PWFSMBQ
SFDU SFDU PWFSMBQ@MFGUNBY SFDU<*/%&9@-&'5> SFDU<*/%&9@-&'5> PWFSMBQ@SJHIUNJO SFDU<*/%&9@3*()5> SFDU<*/%&9@3*()5> PWFSMBQ@UPQNBY SFDU<*/%&9@501> SFDU<*/%&9@501> PWFSMBQ@CPUUPNNJO SFDU<*/%&9@#0550.> SFDU<*/%&9@#0550.> PWFSMBQ@MBOE <PWFSMBQ@MFGU PWFSMBQ@UPQ PWFSMBQ@SJHIU PWFSMBQ@CPUUPN> VOJPO@MBOE SFDU @MBOE SFDU PWFSMBQ SFUVSOPWFSMBQVOJPO

ޡࠩؔ਺ ֬৴౓ʢ$POpEFODFʣˠ.4&ʢ.FBO4RVBSFE&SSPSʣ ࠲ඪʢ3FHJPOʣˠ4NPPUI--PTT ʢ'BTU3$//IUUQTBSYJWPSHBCTʣ

def _smooth_l1_loss(x): with tf.name_scope("smooth_l1"): abs_x = tf.abs(x) less_mask = tf.cast(abs_x
< 1.0, tf.float32) return less_mask * (0.5 * tf.square(x)) + (1.0 - less_mask) * (abs_x - 0.5) def _loc_loss(logits, groundtruth): offset = logits - groundtruth return _smooth_l1_loss(offset)

MSEʢMean Squared Errorʣ Smooth L1 Loss

ֶशʹࣦഊʜʜ

1PTJUJWF4BNQMFͷׂ߹ 841 29x29 225 15x15 64 8x8 1 4
2x2 16 4x4 1,151

)BSE/FHBUJWF.JOJOH

1PTJUJWFαϯϓϧʹՃ͑ͯɺޡࠩͷେ͖ͳαϯϓϧΛબ୒

/&("5*7&@$06/5@3"5& EFG@DBMD@MPTT@XJUI@IBSE@OFHBUJWF@NJOJOH MPTTFT QPTJUJWF@NBTL QPTJUJWF@DPVOU OFHBUJWF@DPVOUQPTJUJWF@DPVOU /&("5*7&@$06/5@3"5& CBUDI@TJ[FMPTTFTHFU@TIBQF
<>WBMVF QPTJUJWF@MPTTFTMPTTFT QPTJUJWF@NBTL OFHBUJWF@MPTTFTMPTTFTQPTJUJWF@MPTTFT UPQ@OFHBUJWF@MPTTFT @UGOOUPQ@L UGSFTIBQF OFHBUJWF@MPTTFT <CBUDI@TJ[F > LUGDBTU OFHBUJWF@DPVOU UGJOU MPTT UGSFEVDF@TVN QPTJUJWF@MPTTFTQPTJUJWF@DPVOU UGSFEVDF@TVN UPQ@OFHBUJWF@MPTTFTOFHBUJWF@DPVOU SFUVSOMPTT

ֶशʹࣦഊʜʜ

إ͸্ख͘ݕग़Ͱ͖͍ͯΔ͕ ࠲ඪ͕ਖ਼͘͠ͳ͍

EFGBVMU#PY ʢ"ODIPSʣ

EFGBVMUCPYͷಋೖ left : 0.5000 top : 0.2500 right :
0.7500 bottom : 0.5000 0.0000 : left 0.5000 : top 0.2500 : right 0.7500 : bottom

EFGBVMUCPYͱ࠲ඪ left : 0.5000 top : 0.2500 right :
0.7500 bottom : 0.5000 0.0000 : left 0.5000 : top 0.2500 : right 0.7500 : bottom left : 0.6036 top : 0.2854 right : 0.8220 bottom : 0.5236 0.1092 : left 0.4094 : top 0.3209 : right 0.6247 : bottom

EFGBVMUCPYͱ࠲ඪͷ ࠩʢP⒎TFUʣΛઃఆ left : 0.6036 top : 0.2854 right
: 0.8220 bottom : 0.5236 left : 0.1036 top : 0.0354 right : 0.0720 bottom : 0.0236 0.1092 : left -0.0906 : top 0.0709 : right -0.1253 : bottom 0.1092 : left 0.4094 : top 0.3209 : right 0.6247 : bottom

ग़ྗ ʹҙຯ͕ੜ·ΕΔ

׆ੑԽؔ਺ͷมߋʢग़ྗ૚ʣ left : 0.1036 top : 0.0354 right : 0.0720
bottom : 0.0236 0.1092 : left -0.0906 : top 0.0709 : right -0.1253 : bottom

׆ੑԽؔ਺ͷมߋʢग़ྗ૚ʣ Sigmoid Hyperbolic Tangent

ֶशɾ܇࿅ εςοϓʢόοναΠζ Ϛϧν(16ʣ  ࠷దԽΞϧΰϦζϜ"EBN0QUJNJ[FS ֶश཰ ݮਰ TUFQ

؀ڥʢ͘͞ΒͷߴՐྗίϯϐϡʔςΟϯάʣ $169FPO$PSFʷ .FNPSZ(# 44%(# (F'PSDF(595*5"/9ʢ1BTDBMΞʔΩςΫνϟʣ(#ʷ (F'PSDF(595Jʢ1BTDBMΞʔΩςΫνϟʣ(#ʷ

ޡࠩͷมԽ

ࠓޙͷ՝୊

ΞεϖΫτൺ͕େ͖͍ը૾΁ͷରԠ

ΞεϖΫτൺͷҟͳΔEFGBVMUCPYͷಋೖ

ٕज़ॻయ̏ https://techbookfest.org/event/tbf03/ circle/5686003050217472 ೔࣌ɿ2017೥10݄22೔ʢ೔ʣ  ɹɹɹ11࣌ʙ17࣌ ৔ॴɿΞΩόɾεΫΤΞ ओ࠵ɿTechbooster / ୡਓग़൛ձ Illustration:
ࠜઇΕ͍

ࢀߟ 4JOHMF4IPU.VMUJ#PY%FUFDUPS  IUUQTBSYJWPSHBCT 44%4JOHMF4IPU.VMUJ#PY%FUFDUPSʢ೔ຊޠ༁ʣ2JJUB  IUUQTRJJUBDPNEFUBJUFNTBFDFGDBB 5FOTPS'MPXͰإݕग़ثΛࣗ࡞͢Δ͗͢ΌʔΜϝϞ  IUUQNFNPTVHZBODPNFOUSZ 5FOTPS'MPXͰ෺ମྖҬ༧ଌʢ3FHJPO1SPQPTBMʣΛࢼͯ͠ΈΔ  IUUQXPSLQJMFTDPNUFOTPSqPXTTESFHJPOQSPQPTBM

ࢀߟ 'BTU3$//  IUUQTBSYJWPSHBCT 'BTUFS3$//5PXBSET3FBM5*NF0CKFDU%FUFDUJPOXJUI3FHJPO1SPQPTBM/FUXPSLT  IUUQTBSYJWPSHBCT :PV0OMZ-PPL0ODF6OJpFE 3FBM5JNF0CKFDU%FUFDUJPO  IUUQTBSYJWPSHBCT +BDDBSE*OEFY  IUUQTFOXJLJQFEJBPSHXJLJ+BDDBSE@JOEFY

C-LIS CO., LTD. ຊࢿྉ͸ɺ༗ݶձࣾγʔϦεͷஶ࡞෺Ͱ͢ɻܝࡌ͞Ε͍ͯΔΠϥετ͸ɺಛʹهࡌ͕ͳ͍৔߹͸ࠜઇΕ͍ͷஶ࡞෺Ͱ͢ɻ ຊࢿྉͷશ෦ɺ·ͨ͸Ұ෦ʹ͍ͭͯɺஶ࡞ऀ͔ΒจॻʹΑΔڐ୚Λಘͣʹෳ੡͢Δ͜ͱ͸ې͡ΒΕ͍ͯ·͢ɻ 5IF"OESPJE4UVEJPJDPOJTSFQSPEVDFEPSNPEJpFEGSPNXPSLDSFBUFEBOETIBSFECZ(PPHMFBOEVTFEBDDPSEJOHUPUFSNTEFTDSJCFEJOUIF$SFBUJWF$PNNPOT"UUSJCVUJPO-JDFOTF ֤੡඼໊ɾϒϥϯυ໊ɺձ໊ࣾͳͲ͸ɺҰൠʹ֤ࣾͷ঎ඪ·ͨ͸ొ࿥঎ඪͰ͢ɻຊࢿྉதͰ͸ɺɺɺäΛׂѪ͍ͯ͠·͢ɻ 5IF"OESPJESPCPUJTSFQSPEVDFEPSNPEJpFEGSPNXPSLDSFBUFEBOETIBSFECZ(PPHMFBOEVTFEBDDPSEJOHUPUFSNTEFTDSJCFEJOUIF$SFBUJWF$PNNPOT"UUSJCVUJPO-JDFOTF https://speakerdeck.com/keiji/devfest-tokyo-2017-tensorﬂow

DevFest Tokyo 2017 TensorFlow

DevFest Tokyo 2017 TensorFlow

More Decks by ARIYAMA Keiji

Other Decks in Technology

Featured

Transcript