使用tensorflow识别乐高积木？
Using tensorflow to identify lego bricks?
读过 this article 关于一个人使用 tensorflow 将黄瓜分成九种不同的 类 我想知道这种类型的过程是否可以应用于大量 类 .我的想法是用它来识别乐高零件。
目前，像 Bricklink 这样的网站描述的内容超过 40,000 different parts，因此它与黄瓜示例有点不同，但我想知道它听起来是否合适。没有简单的方法可以为每个部分获取数百张图片，但以下过程听起来是否可行：
- 拍摄一部分；
- 尝试使用 tensorflow 识别该部分；
- 如果它没有识别出正确的部分，拍摄更多照片并将它们输入神经网络；
- 继续下一部分。
这样，每次我们遇到新作品时，我们都会 "teach" 网络参考，以便下次更好地识别它。像这样，经过人工监控的数百次迭代，我们可以想象 tensorflow 能够识别零件吗？至少最常见的？
我的问题可能听起来很愚蠢，但我不喜欢神经网络，所以欢迎任何建议。目前我还没有找到任何方法来根据图片识别乐高零件，这 "cucumber example" 听起来很有希望，所以我正在寻找一些反馈。
谢谢。
我已经开始使用 IBM Watson 的 Visual Recognition。
我有六种不同的积木要在运输带背景上识别。
我实际上在考虑 tensorflow，因为我可以在本地使用它运行。
代码实验室：TensorFlow for Poets，几乎准确描述了您想要实现的目标，
有关 Watson 版本的演示：
https://www.ibm.com/developerworks/community/blogs/ibmandgoogle/entry/Lego_bricks_recognition_with_Watosn_lego_and_raspberry_pi?lang=en
您可以阅读 Jacques Mattheij 的作品，他实际上在 https://keras.io/ 上使用了定制版的 Xception1 运行。
简介是Sorting 2 Metric Tons of Lego.
在Sorting 2 Tons of Lego, The software Side中您可以阅读：
The hard challenge to deal with next was to get a training set large
enough to make working with 1000+ classes possible. At first this
seemed like an insurmountable problem. I could not figure out how to
make enough images and to label them by hand in acceptable time, even
the most optimistic calculations had me working for 6 months or longer
full-time in order to make a data set that would allow the machine to
work with many classes of parts rather than just a couple.
In the end the solution was staring me in the face for at least a week
before I finally clued in: it doesn’t matter. All that matters is that
the machine labels its own images most of the time and then all I need
to do is correct its mistakes. As it gets better there will be fewer
mistakes. This very rapidly expanded the number of training images.
The first day I managed to hand-label about 500 parts. The next day
the machine added 2000 more, with about half of those labeled wrong.
The resulting 2500 parts where the basis for the next round of
training 3 days later, which resulted in 4000 more parts, 90% of which
were labeled right! So I only had to correct some 400 parts, rinse,
repeat… So, by the end of two weeks there was a dataset of 20K images,
all labeled correctly.
This is far from enough, some classes are severely under-represented
so I need to increase the number of images for those, perhaps I’ll
just run a single batch consisting of nothing but those parts through
the machine. No need for corrections, they’ll all be labeled
identically.
最近的更新是 Sorting 2 Tons of Lego, Many Questions, Results。
1CHOLLET, François. Xception: Deep Learning with Depthwise Separable Convolutions. arXiv preprint arXiv:1610.02357, 2016.
读过 this article 关于一个人使用 tensorflow 将黄瓜分成九种不同的 类 我想知道这种类型的过程是否可以应用于大量 类 .我的想法是用它来识别乐高零件。
目前，像 Bricklink 这样的网站描述的内容超过 40,000 different parts，因此它与黄瓜示例有点不同，但我想知道它听起来是否合适。没有简单的方法可以为每个部分获取数百张图片，但以下过程听起来是否可行：
- 拍摄一部分；
- 尝试使用 tensorflow 识别该部分；
- 如果它没有识别出正确的部分，拍摄更多照片并将它们输入神经网络；
- 继续下一部分。
这样，每次我们遇到新作品时，我们都会 "teach" 网络参考，以便下次更好地识别它。像这样，经过人工监控的数百次迭代，我们可以想象 tensorflow 能够识别零件吗？至少最常见的？
我的问题可能听起来很愚蠢，但我不喜欢神经网络，所以欢迎任何建议。目前我还没有找到任何方法来根据图片识别乐高零件，这 "cucumber example" 听起来很有希望，所以我正在寻找一些反馈。
谢谢。
我已经开始使用 IBM Watson 的 Visual Recognition。 我有六种不同的积木要在运输带背景上识别。
我实际上在考虑 tensorflow，因为我可以在本地使用它运行。
代码实验室：TensorFlow for Poets，几乎准确描述了您想要实现的目标，
有关 Watson 版本的演示： https://www.ibm.com/developerworks/community/blogs/ibmandgoogle/entry/Lego_bricks_recognition_with_Watosn_lego_and_raspberry_pi?lang=en
您可以阅读 Jacques Mattheij 的作品，他实际上在 https://keras.io/ 上使用了定制版的 Xception1 运行。
简介是Sorting 2 Metric Tons of Lego.
在Sorting 2 Tons of Lego, The software Side中您可以阅读：
The hard challenge to deal with next was to get a training set large enough to make working with 1000+ classes possible. At first this seemed like an insurmountable problem. I could not figure out how to make enough images and to label them by hand in acceptable time, even the most optimistic calculations had me working for 6 months or longer full-time in order to make a data set that would allow the machine to work with many classes of parts rather than just a couple.
In the end the solution was staring me in the face for at least a week before I finally clued in: it doesn’t matter. All that matters is that the machine labels its own images most of the time and then all I need to do is correct its mistakes. As it gets better there will be fewer mistakes. This very rapidly expanded the number of training images. The first day I managed to hand-label about 500 parts. The next day the machine added 2000 more, with about half of those labeled wrong. The resulting 2500 parts where the basis for the next round of training 3 days later, which resulted in 4000 more parts, 90% of which were labeled right! So I only had to correct some 400 parts, rinse, repeat… So, by the end of two weeks there was a dataset of 20K images, all labeled correctly.
This is far from enough, some classes are severely under-represented so I need to increase the number of images for those, perhaps I’ll just run a single batch consisting of nothing but those parts through the machine. No need for corrections, they’ll all be labeled identically.
最近的更新是 Sorting 2 Tons of Lego, Many Questions, Results。
1CHOLLET, François. Xception: Deep Learning with Depthwise Separable Convolutions. arXiv preprint arXiv:1610.02357, 2016.