Share
Embed
コンピュータビジョンを支える深層学習技術の新潮流
コンピュータビジョン最新技術から、AutoMLやモデルコンパイラ、推論チップなど、開発・運用周りのトレンドをご紹介
Masaki Samejima
February 14, 2019
Transcript
44 slides
-
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. Masaki Samejima Machine Learning Solutions Architect, Amazon Web Services Japan. 2019.2.14 ؝ٝؾُ٦ةؽآّٝ佄ִ 帾㾴㷕统䪮遭ך倜惐崧 Developers Summit 2019 -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. Agenda • ؝ٝؾُ٦ةؽآّٝך⹛ぢ • ؝ٝؾُ٦ةؽآّٝך涪䪮遭 • ؝ٝؾُ٦ةؽآّٝך麊欽䪮遭 • ؝ٝؾُ٦ةؽآّٝ佄ִع٦سؐؑ،ך㾜 -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. ؝ٝؾُ٦ةؽآّٝך⹛ぢ -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. ؝ٝؾُ٦ةؽآّٝהכ • ➂כ歗⫷ַ㢳ֻך䞔㜠䖤ֿהָדֹ • ؝ٝؾُ٦ةחず圫ח➂ך鋔鋙䭯ׇ״ֲהׅ 䪮遭ָ؝ٝؾُ٦ةؽآّٝ Demographic Data Facial Landmarks Sentiment Expressed Image Quality General Attributes 겣ⴓ匿ך⢽ ➂כ겣鋅ג邌䞔 זו铣《 -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. ؝ٝؾُ٦ةؽآّٝחֶֽ帾㾴㷕统ך娖〷 2012 歗⫷钠陎 SuperVision[1] ILSVRC2012⮚ [1] A. Krizhevsky, et al., Imagenet classification with deep convolutional neural networks, NIPS 2012. [2] R Girshick, et al., Rich feature hierarchies for accurate object detection and semantic segmentation, CVPR 2014. [3] I.J. Goodfellow, et al., Generative Adversarial Nets, NIPS 2014. [4] V. Badrinarayanan, et al, SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentation. PAMI 2017 2014 暟⡤嗚⳿ R-CNN[2]ָPascal VOCד礵䏝ぢ♳ 歗⫷欰䧭 侯㼎涸欰䧭طح زٙ٦ؙGAN[3] إًؚٝذ٦ءّٝ ؾؙإٕ⽃⡘ךٓك ؚٔٝSegNet[4] 2015 -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. 굲鬨涸ז䚍腉ぢ♳歗⫷钠陎 https://gluon-cv.mxnet.io/model_zoo/classification.html senet_154 resnet_v1d resnet_v1c resnet_v1b resnet_v1 densenet darknet VGG resnet_v2 mobilenet mobilenetv2 0.80 0.75 0.70 Accuracy 1000 2000 䱿锷鸞䏝 #sample/sec. 3000 4000 • ImageNetח㼎׃ג80%ך礵䏝 • V100 GPUד侧⼪卐ך䱿锷鸞䏝 ⰼך㣐ֹׁכًٌٔ؟؎ؤ -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. 굲鬨涸ז䚍腉ぢ♳暟⡤嗚⳿ https://gluon-cv.mxnet.io/model_zoo/detection.html mAP 10 100 䱿锷鸞䏝 #sample/sec. ⰼך㣐ֹׁכ ًٌٔ؟؎ؤ 40 35 30 yolo3 faster_rcnn ssd • 嗚⳿걄㚖ָ姻鍑ה鵚ְ(IoUָ㣐ְֹ)暟⡤ חאְגծmAP礵䏝ָ30-40%玎䏝 • 鸞䏝דכ굲鬨涸חぢ♳ -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. 굲鬨涸ז䚍腉ぢ♳إًؚٝذ٦ءّٝ https://gluon-cv.mxnet.io/model_zoo/segmentation.html 0 10 20 30 40 50 60 70 80 90 100 fcn_resnet101 psp_resnet101 deeplab_resnet101 fcn_resnet101 psp_resnet101 deeplab_resnet101 deeplab_resnet152 COCO ر٦ةإحز VOC ر٦ةإحز IoU 暟⡤׀הח姻׃ֻ陎ⴽדֹ걄㚖ךⶴさָぢ♳ -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. ״넝䏝ז钠陎堣腉 3如⯋ך暟⡤嗚⳿ [1] [1] B. Tekin, et al., Real-Time Seamless Single Shot 6D Object Pose Prediction, CVPR 2018. [2] R. Girdhar, et al., Detect-and-Track: Efficient Pose Estimation in Videos, CVPR 2018. [3] L. Chen, et al., MaskLab: Instance Segmentation by Refining Object Detection with Semantic and Direction Features, CVPR 2018. ⹛歗ך暟⡤嗚⳿٥鷄騊 [2] 暟⡤׀הך إًؚٝذ٦ءّٝ[3] -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. GANך黝欽걄㚖ך䭁㣐 歗⫷ך欰䧭 Noise 歗⫷ך㢌䳔 Text-to-image [3] (and Image-to-text) ظ؎ؤַ歗⫷פ 遹僤歗⫷ַ㖑㔳פ[1] 歗⫷ך넝鍑⫷䏝⻉[2] [1] P. Isola, et al., Image-to-Image Translation with Conditional Adversarial Nets, CVPR 2017. [2] C. Ledig, et al., Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network, CVPR 2017. [3] S. Reed, et al., Generative Adversarial Text to Image Synthesis, ICML 2016. -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. 歗⫷ַ圫ղז؝ٝذؗأزך钠陎 Saliency (岣湡걄㚖) ך嗚⳿ [1] 鋔简倯ぢ嗚⳿ׅ鄲縧ד ⡲䧭׃ر٦ةإحزⵃ欽 [1] N. Liu, et al., PiCANet: Learning Pixel-wise Contextual Attention for Saliency Detection, CVPR 2018. [2] Z. Li, et al., MegaDepth: Learning Single-View Depth Prediction from Internet Photos, CVPR 2018. 歗⫷ך㤴遤䱿㹀 [2] ְז錬䏝ַ乆䕦׃歗⫷ד 㤴遤ֹ䱿㹀׃גر٦ةإحز⡲䧭 -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. 㼰ꆀر٦ةח㼎ׅ䮋䨌 0 2 4 6 8 10 12 14 16 18 20 1 2 3 4 5 6 7 8 9 1011121314151617181920 겣歗⫷ر٦ةإحز 卐侧 歗⫷ID ⴓ겲 㼎韋 ⴓ겲㼎韋כ Ӎ ח⡂גְ [1] O. Vinyals, et al., Matching Networks for One Shot Learning, arXiv:1606.04080 ر٦ةכ䌢ח⼧ⴓח ֽ֮דכזְ • ⦐ղ陎ⴽٌׅرٕ⡲ֿהכ㔭ꨇ • ⡂גְַוֲַדⴻⴽׅ[1] 嫰 鯰 㼎 韋 -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. 侯㼎涸歗⫷ך欰䧭ה䖴 • Deep Learningךٌرٕכ䠐㢩חظ؎ؤח䓲ְ • ػٝتך歗⫷חظ؎ؤ⸇ִהذشؖؠٕהⴻ㹀ׁ⢽ X. Yuan, et al., Adversarial Examples: Attacks and Defenses for Deep Learning, IEEE Trans Neural Netw Learn Syst. 2019. -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. ؝ٝؾُ٦ةؽآّٝך涪䪮遭 -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. ؝ٝؾُ٦ةؽآّٝך涪ך铬겗 • 넝䚍腉ז،ٕ؞ٔؤيך㹋鄲ךꨇ׃ׁ • 帾㾴㷕统ؿٖ٦يٙ٦ؙך✉甧 • ⸬桦涸ז؝٦سך㹋鄲٥رغحؚך㔭ꨇׁ • ر٦ةⳢ椚٥⸇䊨ծٌرٕ锃侭⡲噟ך頾䬐 ع؎ٖكٕ㹋鄲㹋植 ׅؿٖ٦يٙ٦ؙ ONNXח״ؿٖ٦ي ٙ٦ؙךٌرٕ㢌䳔 AutoML Define-by-run ח״ فؚٗٓىؚٝ -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. ع؎ٖكٕ㹋鄲㹋植ׅؿٖ٦يٙ٦ؙ • 剑⯓畭ךٌرٕך㹋鄲װ㷕统幥ٌرٕꂁ䋒 • ِ٦ؠכٌرٕٗ٦س׃גծⱄ㷕统ծ䱿锷ָ〳腉 TensorFlow models TF slim GluonCV ChainerCV PyTorchCV -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. ResNet㹋鄲ך嫰鯰 (Gluon vs MXNet) num_unit = len(units) assert(num_unit == num_stages) data = mx.sym.Variable(name='data') if dtype == 'float32': data = mx.sym.identity(data=data, name='id') else: if dtype == 'float16': data = mx.sym.Cast(data=data, dtype=np.float16) data = mx.sym.BatchNorm(data=data, fix_gamma=True, eps=2e-5, momentum=bn_mom, name='bn_data') (nchannel, height, width) = image_shape if height <= 32: # such as cifar10 body = mx.sym.Convolution(data=data, num_filter=filter_list[0], kernel=(3, 3), stride=(1,1), pad=(1, 1), no_bias=True, name="conv0", workspace=workspace) else: # often expected to be 224 such as imagenet body = mx.sym.Convolution(data=data, num_filter=filter_list[0], kernel=(7, 7), stride=(2,2), pad=(3, 3), no_bias=True, name="conv0", workspace=workspace) body = mx.sym.BatchNorm(data=body, fix_gamma=False, eps=2e-5, momentum=bn_mom, name='bn0') body = mx.sym.Activation(data=body, act_type='relu', name='relu0') body = mx.sym.Pooling(data=body, kernel=(3, 3), stride=(2,2), pad=(1,1), pool_type='max') for i in range(num_stages): body = residual_unit(body, filter_list[i+1], (1 if i==0 else 2, 1 if i==0 else 2), False, name='stage%d_unit%d' % (i + 1, 1), bottle_neck=bottle_neck, workspace=workspace, memonger=memonger) for j in range(units[i]-1): body = residual_unit(body, filter_list[i+1], (1,1), True, name='stage%d_unit%d' % (i + 1, j + 2), bottle_neck=bottle_neck, workspace=workspace, memonger=memonger) bn1 = mx.sym.BatchNorm(data=body, fix_gamma=False, eps=2e-5, momentum=bn_mom, name='bn1') relu1 = mx.sym.Activation(data=bn1, act_type='relu', name='relu1') MXNet 㹋鄲♧鿇 from mxnet.gluon.model_zoo import vision resnet18 = vision.resnet18_v1() Gluon 㹋鄲 -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. ONNX (Open Neural Network Exchange) MXNet Caffe2 PyTorch TF CNTK CoreML Tensor RT NGraph SNPE • ぐ珏ؿٖ٦يٙ٦ؙד圓眠׃ ٌرٕONNXח㢌䳔ծONNX ַⴽךؿٖ٦يٙ٦ؙח㢌䳔 〳腉 • ؔ٦فٝا٦أד֮ծ؝ىُ صذ؍ח״ג涪ׁגְ -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. ONNX ךⰻ鿇圓䧭 Protocol Buffers ךغ؎شٔؿ؋؎ٕ • ؝ٝػؙزד⸬桦涸 • ֮ײفٓحزؿؓ٦يד⹛⡲ • APIח״铣鴥٥剅ֹ鴥ָ〳 腉 Protocol Buffers Graph Operator Tensor, … Operator Definitions ONNX Python API -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. Define-and-run ה Define-by-run • Define-and-run • طحزٙ٦ؙ㹀纏׃גַر٦ةⰅ⸂ׅ倯䒭 • TensorFlow, MXNetָⴱ劍ַ㼪Ⰵ • Define-by-run • طحزٙ٦ؙ㹀纏׃זָر٦ةⰅ⸂ׅ倯䒭 • Chainerָ䱰欽ծך䖓ծPyTorch, TensorFlow, MXNetפה岚⿹ -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. Define-and-run ה Define-by-run ך⢽ Define-and-run Define-by-run def our_function(A, B): C = A + B return C A = Load_Data_A() B = Load_Data_B() result = our_function(A, B) A = placeholder() B = placeholder() C = A + B our_function = compile(inputs=[A, B], outputs =[C]) A = Load_Data_A() B = Load_Data_B() result = our_function(A, B) https://gluon.mxnet.io/chapter07_distributed-learning/hybridize.html Define Run Define, Run -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. Define-by-run ךًٔحز Define-and-run Define-by-run def our_function(A, B): C = A + B return C A = Load_Data_A() B = Load_Data_B() result = our_function(A, B) A = placeholder() B = placeholder() C = A + B our_function = compile(inputs=[A, B], outputs =[C]) A = Load_Data_A() B = Load_Data_B() result = our_function(A, B) ر٦ة䟝㹀׃鏣鎘 ָ䗳銲㢌⻉ׅر٦ ة䪔ְבְ 㢌⻉חさׇ鏣鎘٥رغح ָؚ〳腉 -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. AutoML הכ • 堣唒㷕统㹋欽⻉ׅחכ圫ղזةأָؙ㶷㖈 • ر٦ة鼅ⴽ٥欰䧭暴䗙䬄⳿ٌرٕ鼅䫛, etc. D. Bayor, et al., TFX: A TensorFlow-Based Production-Scale Machine Learning Platform, KDD 2017. -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. AutoML הכ • ꬊ㼔㹺ךח䗳銲זةأؙ荈⹛⻉ׅ AutoML ך灇瑔ָ鹌 • ICML 2014ַ AutoML ךٙ٦ؙءّحف*ָأة٦ز • ⚺銲ז灇瑔زؾحؙ • ع؎ػ٦ػًٓ٦ةך剑黝⻉ 堣唒㷕统חְֶגծ➂䩛ח״鏣㹀ָ䗳銲ז ع؎ػ٦ػًٓ٦ة⸬桦״ֻ䱱稊 • Meta-Learning, Learning to learn 堣唒㷕统דⵃ欽ׁ剑黝⻉䩛岀זוر٦ة ַ㷕统ׅ倯岀㷕统ׅ倯岀㷕统ׅ * https://sites.google.com/site/automlwsicml14/ -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. AutoML ָׅ堣唒㷕统ءأذي 荈⹛דر٦ة ꧊٥侭⪒ 荈⹛ד剑黝ז堣唒㷕统㹋遤 ِ٦ؠכ堣唒㷕统ך،ٕ؞ٔؤير٦ةך侭⪒倯岀 孡חׅ䗳銲ָזְ -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. AutoML ך⢽Amazon Forecast User CSV file 1. S3חر٦ة،حفٗ٦س 2. Forecast חر٦ةإحز ה✮庠،ٕ؞ٔؤيך鏣㹀 3. Forecast ָ✮庠ٌرٕ圓眠 4. ٌرٕח״儗禸ر٦ة ך✮庠זןח〳鋔⻉ ِ٦ؠחהג剑⡚ꣲ䗳銲זֿהכծر٦ةך،حفٗ٦سהⴱ劍鏣㹀 -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. ؝ٝؾُ٦ةؽآّٝך麊欽䪮遭 -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. ؝ٝؾُ٦ةؽآّٝך麊欽ך铬겗 • 㷕统׃ٌرٕד⸬桦״ֻ䱿锷ׅך 橆㞮㺁僒ח圓眠׃ְ • ٌرٕח״䱿锷穠卓ך㧅䔲䚍鐰⣣דֹ זְ Model Server ٌرٕ؝ٝػ؎ٓ Interpretable ML -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. Model Server ٌرٕرفٗ؎ך铬겗 • 堣唒㷕统ٌرٕח㼎׃גծ䱿锷ؙٔؒأزⳢ椚ׅ橆㞮ָ䗳銲 • 䱿锷橆㞮ך圓眠如痥דכծٖ؎ذٝءָ㉏겗חזֿה֮ Model Server ך䕵ⶴ • ⸬桦ך葺ְ䱿锷橆㞮㺁僒ח圓眠〳 • REST/RPCח״؎ٝة٦ؿؑ٦أדؙٓ؎،ٝزַⵃ欽׃װְׅ 堣唒㷕统ٌرٕ Model Server Mobile client Deploy REST/RPC -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. TensorFlow Serving [1] C. Olston, et al., TensorFlow-Serving: Flexible, High-Performance ML Serving, NIPS 2017. • Controller, Synchronizer穗גծ䱿锷欽ךServing job 醱侧欰䧭 • Routerכծ䱿锷ؙٔؒأز「ֽהծServing job ח鷏⥋ -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. MXNet Model Server https://aws.amazon.com/jp/blogs/news/model-server-for-apache-mxnet-v1-0-released/ • ٌرٕ涫ꐮծ䱿锷ؙٔ ؒأزך鷏⥋REST API䕎䒭ד遤ִ • MMS 1.0ד䚍腉ָぢ♳ ׃ծず儗1,000،ؙإأ דⰋגⳢ椚〳腉 MMS 1.0 MMS 0.4 -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. 帾㾴㷕统ٌرٕ؝ٝػ؎ٓ • 帾㾴㷕统ךٌرٕ㹋遤ׅع٦س ؐؑ،ח䘔ׄגٌرٕ㢌䳔׃ծٌ رٕך䱿锷鸞䏝ぢ♳ • ؝ٝػ؎ٕ䖓ךٌرٕכ鯪ꆀזٓٝ ة؎يד㹋遤〳腉דծ؟؎ؤך㣐ֹ ְ帾㾴㷕统ؿٖ٦يٙ٦ָؙ♶銲 • ؝ٝػ؎ٓך⢽ • AWS, SageMaker Neo • Nvidia, TensorRT Raspberry Piך䱿锷儗 ResNet18 Mobilenet 11.5x 2.2x -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. SageMaker Neo / TVM ח״剑黝⻉ • Operator Fusion 怴皾מהתהח׃ג넝鸞⻉ • Data Layout Transformation 4x4ך遤怴皾遤ֲ㜥さכծر٦ة4x4 ךة؎ٕ朐חׅ • Tensor Expression and Schedule Space ⸬桦⻉ׅח⡚ٖكٕז邌植ח㢌䳔ծ怴皾أ؛آُ٦ؚٔٝ • Nested Parallelism with Cooperation ぐأٖحسָر٦ة⼿⸂׃ג《ח遤ֹծⰟ剣ًٌٔחְֶגְֻ • etc… T. Chen, et al., TVM: An Automated End-to-End Optimizing Compiler for Deep Learning, OSDI 2018. -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. TensorRT ח״剑黝⻉ • Layer & Tensor Fusion 醱侧ךٖ؎َח״Ⳣ椚1אךٖ؎َךⳢ椚חתה • FP16 and INT8 Precision Calibration FP32ך➿חFP16װINT8ⵃ欽׃ג؝ٝػؙزח׃ծ怴皾넝鸞⻉ • Kernel Auto-Tuning ثُ٦صׁؚٝؕ٦طٕ荈⹛鼅䫛׃ծ⢽ִלծ殴鴥ך怴皾剑黝⻉ • Dynamic Tensor Memory ًٌٔךⵃ欽幾׃ֿׅהדծًٌٔⶴ䔲גךؔ٦غقحسꤐ • Multi Stream Execution 醱侧ךⰅ⸂أزٔ٦يד⚛חⳢ椚׃ג䱿锷〳腉 https://devblogs.nvidia.com/tensorrt-3-faster-tensorflow-inference/ -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. Interpretable ML: ⽃秪זٌرٕ ءٝفٕז堣唒㷕统ٌرٕדכծٌرٕ鍑匿׃גծוךر٦ةָ✮ 庠װⴓ겲ח剣⸬ַ濼ֿהָדֹ⢽: 寸㹀加ծSVMծGBTկ C. Molnar, Interpretable Machine Learning, https://christophm.github.io/interpretable-ml-book/ ➂〡 >900♰ ➂〡 < 900♰ 琎 < 2000 km2 匌❨鿪 㣐ꢻ䏍 琎 > 2000 km2 ❨鿪䏍 ⢽ִלծ寸㹀加ך㜥さծ ♳ךظ٦سךקֲָꅾ銲 דծֿך㜥さծ➂〡ָⴓ 겲ךꅾ銲ז䩛ַָ -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. Interpretable ML for computer vision • 帾㾴㷕统ך״ֲח醱꧟⻉׃ٌرٕדכծٌرٕך湫䱸涸ז鍑匿כ 㔭ꨇկ؝ٝؾُ٦ةؽآّٝך㜥さכծ歗⫷Ⰵ⸂׃ג䮙⹛鍑匿կ • 歗⫷ך♧鿇妀衅ׇׁג姻׃ֻ钠陎דֹלծ婍鿇ⴓכ钠陎 חְֶגꅾ銲ז皘䨽،ٝؕ٦הזׇկ ⯋歗⫷ ،ٝؕ٦ ،ٝؕ٦⟃㢩何㢌׃ג姻׃ֻ钠陎 M.T. Ribeiro, et al., Anchors: High-Precision Model-Agnostic Explanations, AAAI 2018. -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. ؝ٝؾُ٦ةؽآّٝ佄ִ ع٦سؐؑ،ך㾜 -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. 䱿锷欽ثحف • ⡚؝أزד넝鸞ח⹛⡲ׅ䱿锷ثحفך涪 • 㷕统欽ثحفהכ欽鷿ָ殯ז • 㷕统欽כ侧⼧卐ך歗⫷鋅חغحثה׃ג㷕统 • 䱿锷欽כ1卐1卐ךِ٦ؠךؙٔؒأز • 涪邌ׁגְثحفך⢽ • AWS Inferentia • Intel Nervana -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. Machine Learning on FPGA • 넝⸬桦זFPGA⢪欽׃帾㾴㷕统 • AWS F1 instanceדכծ涪خ٦ٕろ Amazon Machine Imageⵃ欽דֹ • 殴鴥صُ٦ٕٓطحزٙ٦ؙך״ֲח粸鵤׃鎘皾ָ䗳銲 זךכLoop tiling ח״剑黝⻉ָ䗳銲 [1] [1] C. Zhang, et al., Optimizing FPGA-based Accelerator Design for Deep Convolutional Neural Networks, FPGA 2015. -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. 堣唒㷕统ؒحآرغ؎أ • طحزٙ٦ؙ♶銲דծ⡚ٖ؎ ذٝءד䱿锷ָ〳腉 • 㼭㘗ךⰻ詿GPUד帾㾴㷕统 ך䱿锷ָ〳腉ח -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. 堣唒㷕统ؒحآرغ؎أ -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. תה -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. תה • ؝ٝؾُ٦ةؽآّٝך灇瑔כծ礵䏝٥鸞䏝ךぢ♳ֽדזֻծ ،فٔ؛٦ءّٝך䌴䎢ָגְկ • 帾㾴㷕统،فٔ؛٦ءّٝך涪٥麊欽ך⸬桦⻉ָ岣湡ׁגְ կAutoMLהְ荈⹛⻉חꟼׅ灇瑔遤ծ➙䖓כAIך 孖⚺⻉ָ鹌կ • 帾㾴㷕统ח״؝ٝؾُ٦ةؽآّٝծ䱿锷㹋⚅歲ח䘔欽ׅ ؿؑ٦ؤחկ䱿锷ثحفծؒحآرغ؎أָ➙䖓崞鬨կ -
© 2019, Amazon Web Services, Inc. or its Affiliates. All
rights reserved. إحءّٝך׀䠐鋅ծ׀䠬䟝ֶ耀ְַׇֻׁ https://amzn.to/aws_dev