{"id":4285,"date":"2017-08-26T01:00:00","date_gmt":"2017-08-25T17:00:00","guid":{"rendered":"http:\/\/swarma.org\/%swarma-paper%\/"},"modified":"2018-12-05T21:59:45","modified_gmt":"2018-12-05T13:59:45","slug":"%e8%b5%84%e6%ba%90-%e6%b7%b1%e5%ba%a6%e5%ad%a6%e4%b9%a0%e4%b8%ad%e7%9a%84%e4%bb%a3%e7%a0%81%e8%b5%84%e6%ba%90%e5%ba%93%ef%bc%88%e4%b8%80%ef%bc%89%e5%9b%be%e5%83%8f%e5%a4%84%e7%90%86%e7%af%87","status":"publish","type":"post","link":"https:\/\/swarma.org\/?p=4285","title":{"rendered":"\u8d44\u6e90 | \u6df1\u5ea6\u5b66\u4e60\u4e2d\u7684\u4ee3\u7801\u8d44\u6e90\u5e93\uff08\u4e00\uff09\u56fe\u50cf\u5904\u7406\u7bc7"},"content":{"rendered":"
\n
\n
\"\u96c6\u667a\"<\/div>\n
\n
\n
\n
\n
\n
\n
\n
\n
\n
\n
<\/section>\n<\/section>\n
\n
\n
\n

<\/span><\/p>\n

\u672c\u6587\u662f\u96c6\u667a\u5c0f\u4ed9\u5973\u4e3a\u5927\u5bb6\u6574\u7406\u7684\u4ee3\u7801\u8d44\u6e90\u5e93\u2014\u56fe\u50cf\u5904\u7406\u7bc7\uff0c\u6536\u96c6\u4e86\u5927\u91cf\u6df1\u5ea6\u5b66\u4e60\u9879\u76ee\u56fe\u50cf\u5904\u7406\u9886\u57df\u7684\u4ee3\u7801\u94fe\u63a5\u3002\u5305\u62ec\u56fe\u50cf\u8bc6\u522b\uff0c\u56fe\u50cf\u751f\u6210\uff0c\u770b\u56fe\u8bf4\u8bdd\u7b49\u7b49\u65b9\u5411\u7684\u4ee3\u7801,<\/strong>\u6240\u6709\u4ee3\u7801\u5747\u6309\u7167\u6240\u5c5e\u6280\u672f\u9886\u57df\u9644\u6709\u94fe\u63a5\u3002\u5927\u5bb6\u540c\u65f6\u53ef\u6233\u77e5\u4e4e\u4e2d\u7684\u94fe\u63a5\u7d22\u5f15\uff0c\u76f4\u63a5\u8fdb\u5165\u76f8\u5e94\u94fe\u63a5\u3002<\/span><\/p>\n

\u5f20\u6c5f\u77e5\u4e4e\u4e13\u680f\uff1a<\/span><\/p>\n

https:\/\/zhuanlan.zhihu.com\/p\/28601032?utm_source=wechat_session&utm_medium=social<\/span><\/p>\n

\uff08\u6216\u641c\u7d22\u4eba\u5de5\u667a\u80fd\u7b14\u8bb0\u4e13\u680f\uff09<\/span><\/p>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n

\n
\n
\n

1\u3001 \u56fe\u50cf\u751f\u6210<\/strong> <\/span><\/p>\n

1.1 \u7ed8\u753b\u98ce\u683c\u5230\u56fe\u7247\u7684\u8f6c\u6362\uff1aNeural Style <\/p>\n

1.2 \u56fe\u50cf\u7c7b\u6bd4\u8f6c\u6362\uff1aimage-analogies <\/p>\n

1.3 \u6839\u636e\u6d82\u9e26\u751f\u6210\u56fe\u7247\uff1aNeural Doodle <\/p>\n

1.4 \u5339\u6839\u636e\u6d82\u9e26\u7c7b\u6bd4\u56fe\u7247\uff1aSketchy <\/p>\n

1.5 \u6839\u636e\u56fe\u7247\u751f\u6210\u94c5\u7b14\u753b\uff1aPencil <\/p>\n

1.6 \u624b\u5199\u6587\u5b57\u6a21\u62df\uff1arnnlib <\/p>\n

1.7 \u8f6c\u6362\u98ce\u666f\u56fe\u7247\uff1aTransient Attributes for High-Level Understanding and Editing of Outdoor Scenes <\/p>\n

1.8 \u56fe\u7247\u53d8Emojis\u8868\u60c5\uff1aWhat emojis will the Emojini 3000 grant your photos? <\/p>\n

1.9 \u589e\u52a0\u56fe\u7247\u5206\u8fa8\u7387\uff1asrez <\/p>\n

1.10 \u56fe\u7247\u81ea\u52a8\u4e0a\u8272\uff1aColornet <\/p>\n

1.11 \u751f\u6210\u53ef\u7231\u7684\u52a8\u6f2b\u5934\u50cf\uff1aAnimeGAN <\/p>\n

1.12 \u9aa1\u5b50\u53d8\u6591\u9a6c\uff1aCycleGAN and pix2pix in PyTorch <\/p>\n

1.13 \u5f3a\u5927\u7684\u56fe\u50cf\u751f\u6210\u5668\uff1aDiscoGAN in PyTorch <\/p>\n

1.14 \u4f7f\u7528RNN\u751f\u6210\u624b\u5199\u6570\u5b57\uff1aDRAW implmentation <\/p>\n

1.15 \u4f7f\u7528CNN\u6765\u653e\u5927\u56fe\u7247\uff1awaifu2x<\/p>\n

2\u3001 \u770b\u56fe\u8bf4\u8bdd <\/strong><\/span><\/p>\n

2.1 \u6839\u636e\u56fe\u7247\u751f\u6210\u4e00\u6bb5\u63cf\u8ff0\uff1aShow and Tell <\/p>\n

2.2 \u6839\u636e\u56fe\u7247\u8bb2\u6545\u4e8b\uff1aneural-storyteller <\/p>\n

2.3 \u6839\u636e\u56fe\u7247\u5c06\u6545\u4e8b2:NeuralTalk2 <\/p>\n

2.4 \u8bc6\u522b\u56fe\u7247\u4e2d\u7684\u6587\u5b57\uff1aCRNN for image-based sequence recognition<\/p>\n

3\u3001 \u56fe\u50cf\u8bc6\u522b <\/strong><\/span><\/p>\n

3.1 \u7528\u4e8e\u7269\u4f53\u8bc6\u522b\u7684\u5168\u5377\u79ef\u7f51\u7edc\uff1aPyTorch-FCN <\/p>\n

3.2 \u5f15\u5165\u6ce8\u610f\u529b\u7684\u5377\u79ef\u7f51\u7edc\uff1aAttention Transfer <\/p>\n

3.3 \u7269\u4f53\u8bc6\u522b\u5b9e\u4f8b\uff1aDeep-Learning <\/p>\n

3.4 \u7269\u4f53\u8bc6\u522bAPI\uff1aTensorflow Object Detection API <\/p>\n

3.5 \u63a8\u7406\u573a\u666f\u7ed3\u6784\uff1aSfMLearner <\/p>\n

3.6 \u7528\u4e8e\u5206\u8fa8\u8272\u60c5\u56fe\u50cf\u7684open_nsfw <\/p>\n

3.7 \u4eba\u8138\u8bc6\u522b\uff1aOpen Face <\/p>\n

3.8 \u6613\u7528\u4eba\u8138\u8bc6\u522b\uff1aFace_recognition <\/p>\n

3.9 \u5feb\u901f\u4eba\u8138\u8bc6\u522b\uff1aMobileID <\/p>\n

3.10 \u56fe\u50cf\u8bc6\u522b\u6846\u67b61\uff1aAlexNet & VGG Net & GoogleNet & ResNet <\/p>\n

3.11 \u56fe\u50cf\u8bc6\u522b\u6846\u67b62\uff1aResNeXt & RCNN & YOLO & SqueezeNet & SegNet <\/p>\n

3.12 \u9884\u8bad\u7ec3\u7684\u56fe\u50cf\u8bc6\u522b\u6a21\u578b\uff1afunctional-zoo <\/p>\n

3.13 \u9884\u5b9a\u4e49\u7684CNN\u8fc7\u6ee4\u5668\uff1a PyScatWave <\/p>\n

3.14 \u8ba1\u7b97\u56fe\u7247\u4e2d\u7269\u4f53\u7684\u76f8\u4f3c\u5ea6\uff1aConditional Similarity Networks (CSNs) <\/p>\n

3.15 \u91cf\u5b50\u5316\u5b66\u4e2d\u7684\u795e\u7ecf\u4fe1\u606f\u4f20\u9012(?_?;Neural Message Passing for Quantum Chemistry<\/p>\n

4 \u3001\u56fe\u50cf\u7406\u89e3 <\/strong><\/span><\/p>\n

4.1 Visual Question Answering in Pytorch <\/p>\n

4.2 Facebook\u770b\u56fe\u7b54\u9898\uff1aClevr-IEP<\/p>\n<\/section>\n<\/section>\n<\/section>\n


<\/span><\/p>\n

\n
\n
\n
\n
\n
1<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n

\u56fe\u50cf\u751f\u6210<\/span><\/strong><\/span><\/inherit><\/p>\n<\/section>\n

\u7ed8\u753b\u98ce\u683c\u5230\u56fe\u7247\u7684\u8f6c\u6362\uff1aNeural Style<\/span><\/strong><\/p>\n

https:\/\/github.com\/jcjohnson\/neural-style<\/span><\/p>\n

\u8fd9\u4e2a\u9879\u76ee\u662f\u7528 Torch \u5bf9 Leon A. Gatys, Alexander S. Ecker, \u548c Matthias Bethge \u7b49\u4eba\u7684\u8bba\u6587\u201cA Neural Algorithm of Artistic Style\u201d\u7684\u4e00\u4e2a\u5b9e\u73b0\u3002\u8bba\u6587\u4e2d\u63d0\u51fa\u4e00\u79cd\u7b97\u6cd5\uff0c\u7528\u5377\u79ef\u795e\u7ecf\u7f51\u7edc\u5c06\u4e00\u5e45\u56fe\u50cf\u7684\u5185\u5bb9\u4e0e\u53e6\u4e00\u5e45\u56fe\u50cf\u7684\u98ce\u683c\u8fdb\u884c\u7ec4\u5408\u3002<\/span><\/p>\n

\u56fe\u50cf\u7c7b\u6bd4\u8f6c\u6362\uff1aimage-analogies<\/span><\/strong><\/p>\n

https:\/\/github.com\/awentzonline\/image-analogies<\/span><\/p>\n

\u201c\u795e\u7ecf\u56fe\u50cf\u7c7b\u6bd4\u201d\uff08neural image analogies\uff09\u8fd9\u4e2a\u9879\u76ee\u57fa\u672c\u4e0a\u662f A. Hertzmann et. al\uff082001\uff09\u7684\u8bba\u6587\u201cImage Analogies\u201d\u7684\u4e00\u4e2a\u5b9e\u73b0\u3002\u5728\u8fd9\u4e2a\u9879\u76ee\u4e2d\uff0c\u6211\u4eec\u4f7f\u7528\u4e86 VGG16 \u7684\u7279\u5f81\uff0c\u5229\u7528 Chuan Li, Michael Wand (2016) \u7684\u8bba\u6587\u201cCombining Markov Random Fields and Convolutional Neural Networks for Image Synthesis\u201d\u4e2d\u63cf\u8ff0\u7684\u65b9\u6cd5\u8fdb\u884cpatch\u7684\u5339\u914d\u548c\u6df7\u5408\u3002\u521d\u59cb\u4ee3\u7801\u6539\u7f16\u81ea Keras \u7684\u201c\u795e\u7ecf\u98ce\u683c\u8fc1\u79fb\u201d\u793a\u4f8b\u3002<\/span><\/p>\n

\u6839\u636e\u6d82\u9e26\u751f\u6210\u56fe\u7247\uff1aNeural Doodle<\/span><\/strong><\/p>\n

https:\/\/github.com\/alexjc\/neural-doodle<\/span><\/p>\n

\u4f7f\u7528\u6df1\u5ea6\u795e\u7ecf\u7f51\u7edc\u628a\u4f60\u7684\u4e8c\u6d41\u6d82\u9e26\u53d8\u6210\u827a\u672f\u4e00\u822c\u7684\u4f5c\u54c1\uff01\u8fd9\u4e2a\u9879\u76ee\u662f Champandard\uff082016\uff09\u7684\u8bba\u6587 \u201cSemantic Style Transfer and Turning Two-Bit Doodles into Fine Artworks\u201d\u7684\u4e00\u4e2a\u5b9e\u73b0\uff0c\u57fa\u4e8e Chuan Li \u548c Michael Wand\uff082016\uff09\u5728\u8bba\u6587\u201cCombining Markov Random Fields and Convolutional Neural Networks for Image Synthesis\u201d\u4e2d\u63d0\u51fa\u7684 Neural Patches \u7b97\u6cd5\u3002<\/span><\/p>\n

\u6839\u636e\u6d82\u9e26\u7c7b\u6bd4\u56fe\u7247\uff1aSketchy<\/span><\/strong><\/p>\n

https:\/\/github.com\/janesjanes\/sketchy<\/span><\/p>\n

\u8fd9\u4e2a\u9879\u76ee\u53ef\u4ee5\u6839\u636e\u7528\u6237\u624b\u7ed8\u7684\u6d82\u9e26\uff0c\u5339\u914d\u51fa\u7c7b\u4f3c\u7684\u56fe\u7247\u3002<\/span><\/p>\n

\u6839\u636e\u56fe\u7247\u751f\u6210\u94c5\u7b14\u753b\uff1aPencil<\/span><\/strong><\/p>\n

https:\/\/github.com\/fumin\/pencil<\/span><\/p>\n

\u628a\u4e00\u526f\u56fe\u50cf\u53d8\u6210\u94c5\u7b14\u6c34\u7c89\u753b\u3002<\/span><\/p>\n

\u624b\u5199\u6587\u5b57\u6a21\u62df\uff1arnnlib<\/span><\/strong><\/p>\n

https:\/\/github.com\/szcom\/rnnlib<\/span><\/p>\n

\u8fd9\u4e2a\u9879\u76ee\u53ef\u4ee5\u505a\u5230\u624b\u5199\u6587\u5b57\u6a21\u62df\u3002<\/span><\/p>\n

\u8f6c\u6362\u98ce\u666f\u56fe\u7247\uff1aTransient Attributes for High-Level Understanding and Editing of Outdoor Scenes<\/span><\/strong><\/p>\n

http:\/\/transattr.cs.brown.edu<\/span><\/p>\n

\u8fd9\u4e2a\u9879\u76ee\u53ef\u4ee5\u8bc6\u522b\u548c\u7406\u89e3\u56fe\u7247\u4e2d\u7684\u98ce\u666f\uff0c\u5e76\u4e14\u53ef\u4ee5\u6839\u636e\u7528\u6237\u63d0\u51fa\u7684\u6761\u4ef6\uff0c\u5b9a\u5411\u6539\u53d8\u539f\u98ce\u666f\u753b\u4e2d\u7684\u73af\u5883\uff08\u6bd4\u5982more night\uff09<\/span><\/p>\n

\u56fe\u7247\u53d8Emojis\u8868\u60c5\uff1aWhat emojis will the Emojini 3000 grant your photos?<\/span><\/strong><\/p>\n

http:\/\/engineering.curalate.com\/2016\/01\/20\/emojinet.html<\/span><\/p>\n

\u5c06\u7528\u6237\u63d0\u4f9b\u7684\u56fe\u7247\u8f6c\u5316\u6210\u76f8\u5173\u7684\u8868\u60c5\u56fe\u6807<\/span><\/p>\n

\u589e\u52a0\u56fe\u7247\u5206\u8fa8\u7387\uff1asrez<\/span><\/strong><\/p>\n

https:\/\/github.com\/david-gpu\/srez<\/span><\/p>\n

srez\uff08super-resolution through deep learning\uff09\uff0c\u5373\u901a\u8fc7\u6df1\u5ea6\u5b66\u4e60\u5b9e\u73b0\u56fe\u50cf\u8d85\u5206\u8fa8\u7387\u3002\u8fd9\u4e2a\u9879\u76ee\u662f\u5229\u7528\u6df1\u5ea6\u5b66\u4e60\u5c06 16×16 \u7684\u56fe\u50cf\u5206\u8fa8\u7387\u589e\u52a0 4 \u500d\uff0c\u57fa\u4e8e\u7528\u6765\u8bad\u7ec3\u795e\u7ecf\u7f51\u7edc\u7684\u6570\u636e\u96c6\uff0c\u6240\u5f97\u5230\u7684\u56fe\u50cf\u5177\u6709\u9c9c\u660e\u7684\u7279\u5f81\u3002<\/span><\/p>\n

\u56fe\u7247\u81ea\u52a8\u4e0a\u8272\uff1aColornet<\/span><\/strong><\/p>\n

https:\/\/github.com\/pavelgonchar\/colornet<\/span><\/p>\n

Colornet \u662f\u4e00\u4e2a\u7ed9\u7070\u5ea6\u56fe\u50cf\u81ea\u52a8\u4e0a\u8272\u7684\u795e\u7ecf\u7f51\u7edc\u3002<\/span><\/p>\n

\u751f\u6210\u53ef\u7231\u7684\u52a8\u6f2b\u5934\u50cf\uff1aAnimeGAN<\/span><\/strong><\/p>\n

https:\/\/github.com\/jayleicn\/animeGAN<\/span><\/p>\n

\u4f7f\u7528PyTorch\u5b9e\u73b0\u7684GAN\uff0c\u53ef\u4ee5\u81ea\u5b9a\u4e49\u751f\u6210\u6f02\u4eae\u7684\u52a8\u6f2b\u59b9\u5b50\u5934\u50cf\uff0c\u9644\u5e26\u8bad\u7ec3\u6570\u636e\u96c6\u54e6\uff01<\/span><\/p>\n

\u9aa1\u5b50\u53d8\u6591\u9a6c\uff1aCycleGAN and pix2pix in PyTorch<\/span><\/strong><\/p>\n

https:\/\/github.com\/junyanz\/pytorch-CycleGAN-and-pix2pix.git<\/span><\/p>\n

\u56fe\u5230\u56fe\u7684\u7ffb\u8bd1\uff0c\u8457\u540d\u7684 CycleGAN \u4ee5\u53ca pix2pix \u7684PyTorch \u5b9e\u73b0\u3002<\/span><\/p>\n

\u5f3a\u5927\u7684\u56fe\u50cf\u751f\u6210\u5668\uff1aDiscoGAN in PyTorch<\/span><\/strong><\/p>\n

https:\/\/github.com\/carpedm20\/DiscoGAN-pytorch.git<\/span><\/p>\n

\u300aLearning to Discover Cross-Domain Relations with Generative Adversarial Networks\u300b\u7684 PyTorch \u5b9e\u73b0\u3002<\/span><\/p>\n

\u4f7f\u7528RNN\u751f\u6210\u624b\u5199\u6570\u5b57\uff1aDRAW implmentation<\/span><\/strong><\/p>\n

https:\/\/github.com\/skaae\/lasagne-draw<\/span><\/p>\n

\u4f7f\u7528RNN\u751f\u6210\u624b\u5199\u4f53\u6570\u5b57\u3002<\/span><\/p>\n

\u4f7f\u7528CNN\u6765\u653e\u5927\u56fe\u7247\uff1awaifu2x<\/span><\/strong><\/p>\n

https:\/\/github.com\/nagadomi\/waifu2x<\/span><\/p>\n

\u4f7f\u7528CNN\u6765\u653e\u5927\u56fe\u7247\uff0c\u4e0e\u666e\u901a\u56fe\u7247\u653e\u5927\u4e0d\u540c\u7684\u662f\uff0c\u4f7f\u7528CNN\u201c\u751f\u6210\u201d\u653e\u5927\uff0c\u4f7f\u4f4e\u5206\u8fa8\u7387\u7684\u56fe\u7247\u5728\u653e\u5927\u540e\u4e5f\u4e0d\u4f1a\u51fa\u73b0\u50cf\u7d20\u952f\u9f7f\u3002<\/span><\/p>\n


<\/p>\n

\n
\n
\n
\n
\n
2<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n

\u770b\u56fe\u8bf4\u8bdd<\/span><\/strong><\/span><\/inherit><\/p>\n<\/section>\n

\u6839\u636e\u56fe\u7247\u751f\u6210\u4e00\u6bb5\u63cf\u8ff0\uff1aShow and Tell<\/span><\/strong><\/p>\n

https:\/\/github.com\/tensorflow\/models\/tree\/master\/im2txt<\/span><\/p>\n

\u8fd9\u662f Oriol Vinyals et. al.\uff082016\uff09\u7684\u8bba\u6587\u201cShow and Tell: Lessons learned from the 2015 MSCOCO Image Captioning Challenge\u201d\u7684\u7528TensorFlow\u5b9e\u73b0\u7684 image-to-text \u56fe\u7247\u8bf4\u660e\u751f\u6210\u6a21\u578b\u3002<\/span><\/p>\n

\u6839\u636e\u56fe\u7247\u8bb2\u6545\u4e8b\uff1aneural-storyteller<\/span><\/strong><\/p>\n

https:\/\/github.com\/ryankiros\/neural-storyteller<\/span><\/p>\n

Neural-storyteller \u662f\u4e00\u4e2a\u80fd\u591f\u6839\u636e\u56fe\u50cf\u5185\u5bb9\u751f\u6210\u4e00\u4e2a\u5c0f\u6545\u4e8b\u7684\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\u3002\u8fd9\u4e2a GitHub \u5e93\u91cc\u5305\u542b\u4e86\u4f7f\u7528\u4efb\u610f\u56fe\u50cf\u751f\u6210\u6545\u4e8b\u7684\u4ee3\u7801\uff0c\u4ee5\u53ca\u7528\u4e8e\u8bad\u7ec3\u65b0\u6a21\u578b\u7684\u8bf4\u660e\u3002<\/span><\/p>\n

\u6839\u636e\u56fe\u7247\u5c06\u6545\u4e8b2:NeuralTalk2<\/span><\/strong><\/p>\n

https:\/\/github.com\/karpathy\/neuraltalk2<\/span><\/p>\n

\u5faa\u73af\u795e\u7ecf\u7f51\u7edc\uff08RNN\uff09\u53ef\u4ee5\u7528\u4e8e\u7ed9\u56fe\u50cf\u53d6\u6807\u9898\u3002NeuralTalk2 \u6bd4\u539f\u59cb\u7248\u672c\u7684 NeuralTalk \u66f4\u5feb\u800c\u4e14\u6027\u80fd\u66f4\u597d\u3002\u4e0e\u539f\u6765\u7684 NeuralTalk \u76f8\u6bd4\uff0cNeuralTalk2 \u7684\u5b9e\u73b0\u662f\u6279\u91cf\u7684\uff0c\u53ef\u4ee5\u4f7f\u7528 Torch \u5728 GPU\u4e0a\u8fd0\u884c\uff0c\u5e76\u4e14\u652f\u6301 CNN \u5fae\u8c03\u3002\u8fd9\u4e9b\u90fd\u4f7f\u5f97\u8bed\u8a00\u6a21\u578b\uff08~100x\uff09\u7684\u8bad\u7ec3\u901f\u5ea6\u5927\u5927\u52a0\u5feb\uff0c\u4f46\u7531\u4e8e\u6211\u4eec\u8fd8\u6709\u4e00\u4e2a VGGNet\uff0c\u56e0\u6b64\u603b\u4f53\u4e0a\u7684\u63d0\u5347\u6ca1\u6709\u5f88\u591a\u3002\u4f46\u662f\u8fd9\u4ecd\u7136\u662f\u4e2a\u597d\u6a21\u578b\uff0c\u53ef\u4ee5\u5728 2~3 \u5929\u91cc\u8bad\u7ec3\u597d\uff0c\u800c\u4e14\u8868\u73b0\u51fa\u7684\u6027\u80fd\u975e\u5e38\u597d\u3002<\/span><\/p>\n

\u8bc6\u522b\u56fe\u7247\u4e2d\u7684\u6587\u5b57\uff1aCRNN for image-based sequence recognition<\/span><\/strong><\/p>\n

https:\/\/github.com\/bgshih\/crnn.git<\/span><\/p>\n

\u8fd9\u4e2a\u662f Convolutional Recurrent Neural Network (CRNN) \u7684 PyTorch \u5b9e\u73b0\u3002CRNN \u7531\u4e00\u4e9bCNN\uff0cRNN\u548cCTC\u7ec4\u6210\uff0c\u5e38\u7528\u4e8e\u57fa\u4e8e\u56fe\u50cf\u7684\u5e8f\u5217\u8bc6\u522b\u4efb\u52a1\uff0c\u4f8b\u5982\u573a\u666f\u6587\u672c\u8bc6\u522b\u548cOCR\u3002<\/span><\/p>\n

<\/span><\/p>\n

\n
\n
\n
\n
\n
3<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n

\u56fe\u50cf\u8bc6\u522b<\/span><\/strong><\/span><\/inherit><\/p>\n<\/section>\n

\u7528\u4e8e\u7269\u4f53\u8bc6\u522b\u7684\u5168\u5377\u79ef\u7f51\u7edc\uff1aPyTorch-FCN<\/span><\/strong><\/p>\n

https:\/\/github.com\/wkentaro\/pytorch-fcn.git<\/span><\/p>\n

\u4e00\u4e2a\u6027\u80fd\u51fa\u4f17\u7684\u7269\u4f53\u8bc6\u522b\u5168\u5377\u79ef\u795e\u7ecf\u7f51\u7edc\uff0c\u4f7f\u7528PyTorch\u5b9e\u73b0\u3002<\/span><\/p>\n

\u5f15\u5165\u6ce8\u610f\u529b\u7684\u5377\u79ef\u7f51\u7edc\uff1aAttention Transfer<\/span><\/strong><\/p>\n

https:\/\/github.com\/szagoruyko\/attention-transfer.git<\/span><\/p>\n

\u8bba\u6587 “Paying More Attention to Attention: Improving the Performance of Convolutional Neural Networks via Attention Transfer” \u7684PyTorch\u5b9e\u73b0\u3002<\/span><\/p>\n

\u7269\u4f53\u8bc6\u522b\u5b9e\u4f8b\uff1aDeep-Learning<\/span><\/strong><\/p>\n

https:\/\/github.com\/priya-dwivedi\/Deep-Learning\/blob\/master\/Object_Detection_Tensorflow_API.ipynb<\/span><\/p>\n

\u4e00\u4e2a\u57fa\u4e8eIpython Notebook\u7684\u7269\u4f53\u8bc6\u522b\u5b9e\u4f8b\uff0c\u4f7f\u7528\u4e86Tensorflow Object Dectection API<\/span><\/p>\n

\u7269\u4f53\u8bc6\u522bAPI\uff1aTensorflow Object Detection API<\/span><\/strong><\/p>\n

https:\/\/github.com\/tensorflow\/models\/tree\/master\/object_detection<\/span><\/p>\n

Google Tensorflow Object Dectection API \u7684\u5f00\u6e90\u4ee3\u7801\u3002<\/span><\/p>\n

\u63a8\u7406\u573a\u666f\u7ed3\u6784\uff1aSfMLearner<\/span><\/strong><\/p>\n

https:\/\/github.com\/tinghuiz\/SfMLearner<\/span><\/p>\n

\u7528\u5355\u5f20\u56fe\u7247\u63a8\u7406\u573a\u666f\u7ed3\u6784\uff1aUC Berkeley\u63d0\u51fa3D\u666f\u6df1\u8054\u5408\u5b66\u4e60\u65b9\u6cd5<\/span><\/p>\n

\u7528\u4e8e\u5206\u8fa8\u8272\u60c5\u56fe\u50cf\u7684open_nsfw<\/span><\/strong><\/p>\n

https:\/\/github.com\/yahoo\/open_nsfw<\/span><\/p>\n

\u8fd9\u662f\u96c5\u864e\u6784\u5efa\u7684\u7528\u4e8e\u68c0\u6d4b\u56fe\u7247\u662f\u5426\u5305\u542b\u4e0d\u9002\u5b9c\u5de5\u4f5c\u573a\u6240\uff08NSFW\uff09\u5185\u5bb9\u7684\u6df1\u5ea6\u795e\u7ecf\u7f51\u7edc\u9879\u76ee\uff0cGitHub \u5e93\u4e2d\u5305\u542b\u4e86\u7f51\u7edc\u7684 Caffe \u6a21\u578b\u7684\u4ee3\u7801\u3002\u68c0\u6d4b\u5177\u6709\u653b\u51fb\u6027\u6216\u6210\u4eba\u5185\u5bb9\u7684\u56fe\u50cf\u662f\u7814\u7a76\u4eba\u5458\u8fdb\u884c\u4e86\u51e0\u5341\u5e74\u7684\u4e00\u4e2a\u96be\u9898\u3002\u968f\u7740\u8ba1\u7b97\u673a\u89c6\u89c9\u6280\u672f\u548c\u6df1\u5ea6\u5b66\u4e60\u7684\u53d1\u5c55\uff0c\u7b97\u6cd5\u5df2\u7ecf\u6210\u719f\uff0c\u96c5\u864e\u7684\u8fd9\u4e2a\u6a21\u578b\u80fd\u4ee5\u66f4\u9ad8\u7684\u7cbe\u5ea6\u5206\u8fa8\u8272\u60c5\u56fe\u50cf\u3002 \u7531\u4e8e NSFW \u754c\u5b9a\u5176\u5b9e\u662f\u5f88\u4e3b\u89c2\u7684\uff0c\u6709\u7684\u4eba\u53cd\u611f\u7684\u4e1c\u897f\u53ef\u80fd\u5176\u4ed6\u4eba\u5e76\u4e0d\u89c9\u5f97\u5982\u4f55\u3002\u96c5\u864e\u7684\u8fd9\u4e2a\u6df1\u5ea6\u795e\u7ecf\u7f51\u7edc\u53ea\u5173\u6ce8NSFW\u5185\u5bb9\u7684\u4e00\u79cd\u7c7b\u578b\uff0c\u5373\u8272\u60c5\u56fe\u7247\uff0c\u6240\u4ee5\u8be5\u6a21\u578b\u4e0d\u9002\u7528\u4e8e\u68c0\u6d4b\u7d20\u63cf\u3001\u6587\u5b57\u3001\u52a8\u753b\u3001\u66b4\u529b\u56fe\u7247\u7b49\u5185\u5bb9\u3002<\/span><\/p>\n

\u4eba\u8138\u8bc6\u522b\uff1aOpen Face<\/span><\/strong><\/p>\n

https:\/\/github.com\/cmusatyalab\/openface<\/span><\/p>\n

OpenFace \u662f\u4e00\u4e2a\u4f7f\u7528\u6df1\u5ea6\u795e\u7ecf\u7f51\u7edc\uff0c\u7528 Python \u548c Torch \u5b9e\u73b0\u4eba\u8138\u8bc6\u522b\u7684\u9879\u76ee\u3002\u795e\u7ecf\u7f51\u7edc\u6a21\u578b\u57fa\u4e8e Google Florian Schroff \u7b49\u4eba\u7684 CVPR 2015 \u8bba\u6587\u201cFaceNet: A Unified Embedding for Face Recognition and Clustering\u201d \uff0cTorch \u8ba9\u7f51\u7edc\u53ef\u4ee5\u5728 CPU \u6216 CUDA \u4e0a\u8fd0\u884c\u3002<\/span><\/p>\n

\u6613\u7528\u4eba\u8138\u8bc6\u522b\uff1aFace_recognition<\/span><\/strong><\/p>\n

https:\/\/github.com\/ageitgey\/face_recognition#face-recognition<\/span><\/p>\n

\u8fd9\u4e5f\u63d0\u4f9b\u4e86\u4e00\u4e2a\u7b80\u5355\u7684 face_recognition \u547d\u4ee4\u884c\u5de5\u5177\uff0c\u4f60\u53ef\u4ee5\u6253\u5f00\u547d\u4ee4\u884c\u4e2d\u4efb\u610f\u56fe\u50cf\u6587\u4ef6\u5939\uff0c\u8fdb\u884c\u4eba\u8138\u8bc6\u522b\uff01<\/span><\/p>\n

\u5feb\u901f\u4eba\u8138\u8bc6\u522b\uff1aMobileID<\/span><\/strong><\/p>\n

https:\/\/github.com\/liuziwei7\/mobile-id<\/span><\/p>\n

\u636e\u8bf4\u662f\u4e2a\u8d85\u7ea7\u5feb\u901f\u7684\u4eba\u8138\u8bc6\u522b\u7a0b\u5e8f\uff0c\u53ef\u4ee5\u7528\u5728\u624b\u673a\u4e0a<\/span><\/p>\n

\u56fe\u50cf\u8bc6\u522b\u6846\u67b61\uff1aAlexNet & VGG Net & GoogleNet & ResNet<\/span><\/p>\n

AlexNet<\/span><\/strong><\/p>\n

https:\/\/gist.github.com\/JBed\/c2fb3ce8ed299f197eff<\/span><\/p>\n

VGG Ne<\/span><\/strong><\/p>\n

https:\/\/github.com\/fchollet\/keras\/blob\/master\/keras\/applications\/vgg16.py<\/span><\/p>\n

GoogleNet<\/span><\/strong><\/p>\n

https:\/\/github.com\/fchollet\/keras\/blob\/master\/keras\/applications\/inception_v3.py<\/span><\/p>\n

ResNet<\/span><\/strong><\/p>\n

https:\/\/github.com\/fchollet\/keras\/blob\/master\/keras\/applications\/resnet50.py<\/span><\/p>\n

\u56fe\u50cf\u8bc6\u522b\u6846\u67b62\uff1aResNeXt & RCNN & YOLO & SqueezeNet & SegNet<\/span><\/strong><\/p>\n

ResNeXt<\/span><\/strong><\/p>\n

https:\/\/github.com\/titu1994\/Keras-ResNeXt<\/span><\/p>\n

RCNN (\u57fa\u4e8e\u533a\u57df\u7684 CNN)<\/span><\/strong><\/p>\n

https:\/\/github.com\/yhenon\/keras-frcnn<\/span><\/p>\n

YOLO (You Only Look once)<\/span><\/strong><\/p>\n

https:\/\/github.com\/allanzelener\/YAD2K<\/span><\/p>\n

SqueezeNet<\/span><\/strong><\/p>\n

https:\/\/github.com\/rcmalli\/keras-squeezenet<\/span><\/p>\n

SegNet<\/span><\/strong><\/p>\n

https:\/\/github.com\/imlab-uiip\/keras-segnet<\/span><\/p>\n

\u9884\u8bad\u7ec3\u7684\u56fe\u50cf\u8bc6\u522b\u6a21\u578b\uff1afunctional-zoo<\/span><\/strong><\/p>\n

https:\/\/github.com\/szagoruyko\/functional-zoo.git<\/span><\/p>\n

\u7531PyTorch\u548cTensorflow\u5b9e\u73b0\u7684\u5e38\u7528\u56fe\u50cf\u8bc6\u522b\u6a21\u578b\u5305\u542b\u9884\u8bad\u7ec3\u53c2\u6570\u3002<\/span><\/p>\n

\u9884\u5b9a\u4e49\u7684CNN\u8fc7\u6ee4\u5668\uff1a PyScatWave<\/span><\/strong><\/p>\n

https:\/\/github.com\/edouardoyallon\/pyscatwave<\/span><\/p>\n

\u4e00\u5957\u9884\u5b9a\u4e49\u7684filter\uff0c\u7528\u4e8e\u589e\u5f3a\u56fe\u50cf\u8bc6\u522b\u7684\u6548\u679c\u3002<\/span><\/p>\n

\u8ba1\u7b97\u56fe\u7247\u4e2d\u7269\u4f53\u7684\u76f8\u4f3c\u5ea6\uff1aConditional Similarity Networks (CSNs)<\/span><\/strong><\/p>\n

https:\/\/github.com\/andreasveit\/conditional-similarity-networks.git<\/span><\/p>\n

\u300aConditional Similarity Networks\u300b\u7684PyTorch\u5b9e\u73b0\uff0c\u53ef\u4ee5\u6839\u636e\u4e0d\u540c\u7684\u6761\u4ef6\u8ba1\u7b97\u56fe\u7247\u4e2d\u7269\u4f53\u7684\u76f8\u4f3c\u5ea6\u3002<\/span><\/p>\n

\u91cf\u5b50\u5316\u5b66\u4e2d\u7684\u795e\u7ecf\u4fe1\u606f\u4f20\u9012(Neural Message Passing for Quantum Chemistry)<\/span><\/strong><\/p>\n

https:\/\/github.com\/priba\/nmp_qc.git<\/span><\/p>\n

\u8bba\u6587\u300aNeural Message Passing for Quantum Chemistry\u300b\u7684PyTorch\u5b9e\u73b0\uff0c\u8bb2\u7684\u662f\u91cf\u5b50\u5316\u5b66\u91cc\u7684\u795e\u7ecf\u4fe1\u606f\u4f20\u9012\uff01\u542c\u8d77\u6765\u7889\u5821\u4e86\u3002<\/span><\/p>\n

<\/span><\/strong><\/p>\n

\n
\n
\n
\n
\n
4<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n

\u56fe\u50cf\u7406\u89e3<\/span><\/strong><\/span><\/inherit><\/p>\n<\/section>\n

Visual Question Answering in Pytorch<\/span><\/strong><\/p>\n

https:\/\/github.com\/Cadene\/vqa.pytorch.git<\/span><\/p>\n

\u4e00\u4e2aPyTorch\u5b9e\u73b0\u7684\u4f18\u79c0\u89c6\u89c9\u63a8\u7406\u95ee\u7b54\u7cfb\u7edf\uff0c\u662f\u57fa\u4e8e\u8bba\u6587\u300aMUTAN: Multimodal Tucker Fusion for Visual Question Answering\u300b\u5b9e\u73b0\u7684\u3002\u9879\u76ee\u4e2d\u6709\u8be6\u7ec6\u7684\u914d\u7f6e\u4f7f\u7528\u65b9\u6cd5\u8bf4\u660e\u3002<\/span><\/p>\n

Facebook\u770b\u56fe\u7b54\u9898\uff1aClevr-IEP<\/span><\/strong><\/p>\n

https:\/\/github.com\/facebookresearch\/clevr-iep.git<\/span><\/p>\n

Facebook Research \u8bba\u6587\u300aInferring and Executing Programs for Visual Reasoning\u300b\u7684PyTorch\u5b9e\u73b0\uff0c\u8bb2\u7684\u662f\u4e00\u4e2a\u53ef\u4ee5\u57fa\u4e8e\u56fe\u7247\u8fdb\u884c\u5173\u7cfb\u63a8\u7406\u95ee\u7b54\u7684\u7f51\u7edc\u3002<\/span><\/p>\n

\n
\n
\n
\n
\n
\n
\n
\n
<\/section>\n
\n

\u66f4\u591a\u5173\u4e8e\u6df1\u5ea6\u5b66\u4e60\u9879\u76ee\u7684\u4ee3\u7801\u94fe\u63a5\uff0c\u89c6\u9891\u7c7b\u3001\u58f0\u97f3\u7c7b\u3001\u8bed\u8a00\u7c7b\u3001\u6e38\u620f\u7c7b\u3001\u6587\u672c\u7c7b\u7b49\uff0c\u5173\u6ce8\u516c\u4f17\u53f7\u6211\u4eec\u4f1a\u6301\u7eed\u66f4\u65b0\uff0c\u656c\u8bf7\u671f\u5f85……<\/strong><\/span><\/p>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n

\u8f6c\u8f7d\u58f0\u660e<\/strong>\uff1a\u672c\u6587\u8f6c\u8f7d\u81ea\u300c\u96c6\u667aAI\u5b66\u56ed\u300d\uff0c\u641c\u7d22\u300cswarmAI\u300d\u5373\u53ef\u5173\u6ce8\u3002
<\/span><\/p>\n

<\/span><\/p>\n

\u63a8\u8350\u9605\u8bfb<\/span><\/strong><\/span><\/p>\n

\u91cd\u78c5\u7cfb\u5217\u8bfe\uff1a\u706b\u70ac\u4e0a\u7684\u6df1\u5ea6\u5b66\u4e60(\u4e0a)<\/span><\/strong><\/span><\/strong><\/a>
<\/p>\n

AI\u6539\u53d8\u4e86\u79d1\u5b66\u2014\u2014\u300a\u79d1\u5b66\u300b\u6700\u65b0AI\u7279\u520a | \u524d\u6cbf<\/a>
<\/span><\/strong><\/span><\/p>\n

\u6cdb\u5728\u7684AI\u4e0e\u6570\u5b57\u4e0a\u5e1d<\/a>
<\/strong><\/span><\/p>\n

\u4eba\u5de5\u667a\u80fd\u793e\u4f1a\u5b66\uff0d\u672a\u6765\u7684\u65b0\u5174\u5b66\u79d1\uff1f<\/a>
<\/span><\/strong><\/span><\/p>\n

\u4ece\u4e07\u7269\u6709\u7075\u5230AI\u638c\u7ba1\u4e16\u754c<\/a>
<\/span><\/strong><\/span><\/p>\n

\u900f\u89c6\u6df1\u5ea6\u5b66\u4e60\uff0c\u7545\u60f3\u672a\u6765\u5e94\u7528<\/a>
<\/span><\/strong><\/span><\/p>\n

\u63ea\u7740\u81ea\u5df1\u7684\u5934\u53d1\u79bb\u5f00\u5730\u7403<\/a><\/span><\/strong><\/span><\/p>\n


<\/p>\n

\n

\"\u96c6\u667a\"
<\/p>\n

\n
\n
\n
\n
\n
\n

\u96c6\u667aQQ\u7fa4\uff5c292641157
\u5546\u52a1\u5408\u4f5c\uff5czhangqian@swarma.org
\u6295\u7a3f\u8f6c\u8f7d\uff5cwangting@swarma.org<\/span><\/strong><\/p>\n

\n

\u25c6 <\/span>\u25c6 <\/span>\u25c6<\/span><\/span><\/strong><\/strong><\/h1>\n<\/section>\n

\u641c\u7d22\u516c\u4f17\u53f7\uff1a\u96c6\u667a\u4ff1\u4e50\u90e8<\/span><\/strong><\/span><\/p>\n


<\/span><\/strong><\/span><\/p>\n

\u52a0\u5165\u201c\u6ca1\u6709\u56f4\u5899\u7684\u7814\u7a76\u6240\u201d<\/span><\/strong><\/span><\/p>\n

\n

\"\u96c6\u667a\"<\/p>\n<\/section>\n

\u8ba9\u82f9\u679c\u7838\u5f97\u66f4\u731b\u70c8\u4e9b\u5427\uff01<\/span><\/p>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n<\/section>\n


<\/p>\n

\n
\n
\"\u96c6\u667a\"<\/section>\n
\u6233\u9605\u8bfb\u539f\u6587\uff0c\u83b7\u53d6\u5b66\u4e60\u89c6\u9891\uff01<\/section>\n<\/section>\n<\/section>\n
\n

\u59cb\u53d1\u4e8e\u5fae\u4fe1\u516c\u4f17\u53f7\uff1a \u96c6\u667a\u4ff1\u4e50\u90e8 <\/p>\n<\/blockquote>\n<\/div>\n","protected":false},"excerpt":{"rendered":"

\u672c\u6587\u662f\u96c6\u667a\u5c0f\u4ed9\u5973\u4e3a\u5927\u5bb6\u6574\u7406\u7684\u4ee3\u7801\u8d44\u6e90\u5e93\u2014\u56fe\u50cf\u5904\u7406\u7bc7\uff0c\u6536\u96c6\u4e86\u5927\u91cf\u6df1\u5ea6\u5b66\u4e60\u9879\u76ee\u56fe\u50cf\u5904\u7406\u9886\u57df\u7684\u4ee3\u7801\u94fe\u63a5\u3002\u5305\u62ec\u56fe\u50cf\u8bc6\u522b\uff0c\u56fe\u50cf\u751f\u6210\uff0c\u770b\u56fe\u8bf4\u8bdd\u7b49\u7b49\u65b9\u5411\u7684\u4ee3\u7801,\u6240\u6709\u4ee3\u7801\u5747\u6309\u7167\u6240\u5c5e\u6280\u672f\u9886\u57df\u9644\u6709\u94fe\u63a5\u3002\u5927\u5bb6\u540c\u65f6\u53ef\u6233\u77e5\u4e4e\u4e2d\u7684\u94fe\u63a5\u7d22\u5f15\uff0c\u76f4\u63a5\u8fdb\u5165\u76f8\u5e94\u94fe\u63a5\u3002 \u5f20\u6c5f\u77e5\u4e4e\u4e13\u680f\uff1a https:\/\/zhuanlan.zhihu.com…<\/p>\n","protected":false},"author":1,"featured_media":4286,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":[],"categories":[19],"tags":[],"special":[],"_links":{"self":[{"href":"https:\/\/swarma.org\/index.php?rest_route=\/wp\/v2\/posts\/4285"}],"collection":[{"href":"https:\/\/swarma.org\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/swarma.org\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/swarma.org\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/swarma.org\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=4285"}],"version-history":[{"count":1,"href":"https:\/\/swarma.org\/index.php?rest_route=\/wp\/v2\/posts\/4285\/revisions"}],"predecessor-version":[{"id":4295,"href":"https:\/\/swarma.org\/index.php?rest_route=\/wp\/v2\/posts\/4285\/revisions\/4295"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/swarma.org\/index.php?rest_route=\/wp\/v2\/media\/4286"}],"wp:attachment":[{"href":"https:\/\/swarma.org\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=4285"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/swarma.org\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=4285"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/swarma.org\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=4285"},{"taxonomy":"special","embeddable":true,"href":"https:\/\/swarma.org\/index.php?rest_route=%2Fwp%2Fv2%2Fspecial&post=4285"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}