mirror of
https://github.com/LCTT/TranslateProject.git
synced 2024-12-23 21:20:42 +08:00
Translated by qhwdw
This commit is contained in:
qhwdw
parent
23229987e5
commit
35086b2d35
@ -1,86 +0,0 @@
|
||||
Transating by qhwdw
|
||||
# [Google launches TensorFlow-based vision recognition kit for RPi Zero W][26]
|
||||
|
||||
|
||||
|
||||
Google’s $45 “AIY Vision Kit” for the Raspberry Pi Zero W performs TensorFlow-based vision recognition using a “VisionBonnet” board with a Movidius chip.
|
||||
|
||||
Google’s AIY Vision Kit for on-device neural network acceleration follows an earlier [AIY Projects][7] voice/AI kit for the Raspberry Pi that shipped to MagPi subscribers back in May. Like the voice kit and the older Google Cardboard VR viewer, the new AIY Vision Kit has a cardboard enclosure. The kit differs from the [Cloud Vision API][8], which was demo’d in 2015 with a Raspberry Pi based GoPiGo robot, in that it runs entirely on local processing power rather than requiring a cloud connection. The AIY Vision Kit is available now for pre-order at $45, with shipments due in early December.
|
||||
|
||||
|
||||
[][9] [][10]
|
||||
**AIY Vision Kit, fully assembled (left) and Raspberry Pi Zero W**
|
||||
(click images to enlarge)
|
||||
|
||||
|
||||
The kit’s key processing element, aside from the 1GHz ARM11-based Broadcom BCM2836 SoC found on the required [Raspberry Pi Zero W][21] SBC, is Google’s new VisionBonnet RPi accessory board. The VisionBonnet pHAT board uses a Movidius MA2450, a version of the [Movidius Myriad 2 VPU][22] processor. On the VisionBonnet, the processor runs Google’s open source [TensorFlow][23]machine intelligence library for neural networking. The chip enables visual perception processing at up to 30 frames per second.
|
||||
|
||||
The AIY Vision Kit requires a user-supplied RPi Zero W, a [Raspberry Pi Camera v2][11], and a 16GB micro SD card for downloading the Linux-based image. The kit includes the VisionBonnet, an RGB arcade-style button, a piezo speaker, a macro/wide lens kit, and the cardboard enclosure. You also get flex cables, standoffs, a tripod mounting nut, and connecting components.
|
||||
|
||||
|
||||
[][12] [][13]
|
||||
**AIY Vision Kit kit components (left) and VisonBonnet accessory board**
|
||||
(click images to enlarge)
|
||||
|
||||
|
||||
Three neural network models are available. There’s a general-purpose model that can recognize 1,000 common objects, a facial detection model that can also score facial expression on a “joy scale” that ranges from “sad” to “laughing,” and a model that can identify whether the image contains a dog, cat, or human. The 1,000-image model derives from Google’s open source [MobileNets][24], a family of TensorFlow based computer vision models designed for the restricted resources of a mobile or embedded device.
|
||||
|
||||
MobileNet models offer low latency and low power consumption, and are parameterized to meet the resource constraints of different use cases. The models can be built for classification, detection, embeddings, and segmentation, says Google. Earlier this month, Google released a developer preview of a mobile-friendly [TensorFlow Lite][14] library for Android and iOS that is compatible with MobileNets and the Android Neural Networks API.
|
||||
|
||||
|
||||
[][15]
|
||||
**AIY Vision Kit assembly views**
|
||||
(click image to enlarge)
|
||||
|
||||
|
||||
In addition to providing the three models, the AIY Vision Kit provides basic TensorFlow code and a compiler, so users can develop their own models. In addition, Python developers can write new software to customize RGB button colors, piezo element sounds, and 4x GPIO pins on the VisionBonnet that can add additional lights, buttons, or servos. Potential models include recognizing food items, opening a dog door based on visual input, sending a text when your car leaves the driveway, or playing particular music based on facial recognition of a person entering the camera’s viewpoint.
|
||||
|
||||
|
||||
[][16] [][17]
|
||||
**Myriad 2 VPU block diagram (left) and reference board**
|
||||
(click image to enlarge)
|
||||
|
||||
|
||||
The Movidius Myriad 2 processor provides TeraFLOPS of performance within a nominal 1 Watt power envelope. The chip appeared on early Project Tango reference platforms, and is built into the Ubuntu-driven [Fathom][25] neural processing USB stick that Movidius debuted in May 2016, prior to being acquired by Intel. According to Movidius, the Myriad 2 is available “in millions of devices on the market today.”
|
||||
|
||||
**Further information**
|
||||
|
||||
The AIY Vision Kit is available for pre-order from Micro Center at $44.99, with shipments due in early December. More information may be found in the AIY Vision Kit [announcement][18], [Google Blog notice][19], and [Micro Center shopping page][20].
|
||||
|
||||
--------------------------------------------------------------------------------
|
||||
|
||||
via: http://linuxgizmos.com/google-launches-tensorflow-based-vision-recognition-kit-for-rpi-zero-w/
|
||||
|
||||
作者：[ Eric Brown][a]
|
||||
译者：[译者ID](https://github.com/译者ID)
|
||||
校对：[校对者ID](https://github.com/校对者ID)
|
||||
|
||||
本文由 [LCTT](https://github.com/LCTT/TranslateProject) 原创编译，[Linux中国](https://linux.cn/) 荣誉推出
|
||||
|
||||
[a]:http://linuxgizmos.com/google-launches-tensorflow-based-vision-recognition-kit-for-rpi-zero-w/
|
||||
[1]:http://twitter.com/share?url=http://linuxgizmos.com/google-launches-tensorflow-based-vision-recognition-kit-for-rpi-zero-w/&text=Google%20launches%20TensorFlow-based%20vision%20recognition%20kit%20for%20RPi%20Zero%20W%20
|
||||
[2]:https://plus.google.com/share?url=http://linuxgizmos.com/google-launches-tensorflow-based-vision-recognition-kit-for-rpi-zero-w/
|
||||
[3]:http://www.facebook.com/sharer.php?u=http://linuxgizmos.com/google-launches-tensorflow-based-vision-recognition-kit-for-rpi-zero-w/
|
||||
[4]:http://www.linkedin.com/shareArticle?mini=true&url=http://linuxgizmos.com/google-launches-tensorflow-based-vision-recognition-kit-for-rpi-zero-w/
|
||||
[5]:http://reddit.com/submit?url=http://linuxgizmos.com/google-launches-tensorflow-based-vision-recognition-kit-for-rpi-zero-w/&title=Google%20launches%20TensorFlow-based%20vision%20recognition%20kit%20for%20RPi%20Zero%20W
|
||||
[6]:mailto:?subject=Google%20launches%20TensorFlow-based%20vision%20recognition%20kit%20for%20RPi%20Zero%20W&body=%20http://linuxgizmos.com/google-launches-tensorflow-based-vision-recognition-kit-for-rpi-zero-w/
|
||||
[7]:http://linuxgizmos.com/free-raspberry-pi-voice-kit-taps-google-assistant-sdk/
|
||||
[8]:http://linuxgizmos.com/google-releases-cloud-vision-api-with-demo-for-pi-based-robot/
|
||||
[9]:http://linuxgizmos.com/files/google_aiyvisionkit.jpg
|
||||
[10]:http://linuxgizmos.com/files/rpi_zerow.jpg
|
||||
[11]:http://linuxgizmos.com/raspberry-pi-cameras-jump-to-8mp-keep-25-dollar-price/
|
||||
[12]:http://linuxgizmos.com/files/google_aiyvisionkit_pieces.jpg
|
||||
[13]:http://linuxgizmos.com/files/google_visionbonnet.jpg
|
||||
[14]:https://developers.googleblog.com/2017/11/announcing-tensorflow-lite.html
|
||||
[15]:http://linuxgizmos.com/files/google_aiyvisionkit_assembly.jpg
|
||||
[16]:http://linuxgizmos.com/files/movidius_myriad2vpu_block.jpg
|
||||
[17]:http://linuxgizmos.com/files/movidius_myriad2_reference_board.jpg
|
||||
[18]:https://blog.google/topics/machine-learning/introducing-aiy-vision-kit-make-devices-see/
|
||||
[19]:https://developers.googleblog.com/2017/11/introducing-aiy-vision-kit-add-computer.html
|
||||
[20]:http://www.microcenter.com/site/content/Google_AIY.aspx?ekw=aiy&rd=1
|
||||
[21]:http://linuxgizmos.com/raspberry-pi-zero-w-adds-wifi-and-bluetooth-for-only-5-more/
|
||||
[22]:https://www.movidius.com/solutions/vision-processing-unit
|
||||
[23]:https://www.tensorflow.org/
|
||||
[24]:https://research.googleblog.com/2017/06/mobilenets-open-source-models-for.html
|
||||
[25]:http://linuxgizmos.com/usb-stick-brings-neural-computing-functions-to-devices/
|
||||
[26]:http://linuxgizmos.com/google-launches-tensorflow-based-vision-recognition-kit-for-rpi-zero-w/
|
||||
@ -0,0 +1,85 @@
|
||||
# [Google 为树莓派 Zero W 发布了基于TensorFlow 的视觉识别套件][26]
|
||||
|
||||
|
||||
|
||||
Google 发行了一个 45 美元的 “AIY Vision Kit”，它是运行在树莓派 Zero W 上的基于 TensorFlow 的视觉识别开发套件，它使用了一个带 Movidius 芯片的 “VisionBonnet” 板。
|
||||
|
||||
为加速设备上的神经网络，Google 的 AIY 视频套件继承了早期树莓派上运行的 [AIY 项目][7] 的语音/AI 套件，这个型号的树莓派随五月份的 MagPi 杂志一起赠送。与语音套件和老的 Google 硬纸板 VR 查看器一样，这个新的 AIY 视觉套件也使用一个硬纸板包装。这个套件和 [Cloud Vision API][8] 是不一样的，它使用了一个在 2015 年演示过的基于树莓派的 GoPiGo 机器人，它完全在本地的处理能力上运行，而不需要使用一个云端连接。这个 AIY 视觉套件现在可以 45 美元的价格去预订，将在 12 月份发货。
|
||||
|
||||
|
||||
[][9] [][10]
|
||||
**AIY 视觉套件，完整包装（左）和树莓派 Zero W**
|
||||
（点击图片放大）
|
||||
|
||||
|
||||
这个套件的主要处理部分除了所需要的 [树莓派 Zero W][21] 单片机之外 —— 一个基于 ARM11 的 1 GHz 的 Broadcom BCM2836 片上系统，另外的就是 Google 最新的 VisionBonnet RPi 附件板。这个 VisionBonnet pHAT 附件板使用了一个 Movidius MA2450，它是 [Movidius Myriad 2 VPU][22] 版的处理器。在 VisionBonnet 上，处理器为神经网络运行了 Google 的开源机器学习库 [TensorFlow][23]。因为这个芯片，便得视觉处理的速度最高达每秒 30 帧。
|
||||
|
||||
这个 AIY 视觉套件要求用户提供一个树莓派 Zero W、一个 [树莓派摄像机 v2][11]、以及一个 16GB 的 micro SD 卡，它用来下载基于 Linux 的 OS 镜像。这个套件包含了 VisionBonnet、一个 RGB 街机风格的按钮、一个压电扬声器、一个广角镜头套件、以及一个包裹它们的硬纸板。还有一些就是线缆、支架、安装螺母、以及连接部件。
|
||||
|
||||
|
||||
[][12] [][13]
|
||||
**AIY 视觉套件组件（左）和 VisonBonnet 附件板**
|
||||
(点击图片放大)
|
||||
|
||||
|
||||
有三个可用的神经网络模型。一个是通用的模型，它可以识别常见的 1,000 个东西，一个是面部检测模型，它可以对 “快乐程度” 进行评分，从 “悲伤” 到 “大笑”，还有一个模型可以用来辨别图像内容是狗、猫、还是人。这个 1,000-image 模型源自 Google 的开源 [MobileNets][24]，它是基于 TensorFlow 家族的计算机视觉模型，它设计用于资源受限的移动或者嵌入式设备。
|
||||
|
||||
MobileNet 模型是低延时、低功耗，和参数化的，以满足资源受限的不同使用案例。Google 说，这个模型可以用于构建分类、检测、嵌入、以及分隔。在本月的早些时候，Google 发布了一个开发者预览版，它是一个对 Android 和 iOS 移动设备友好的 [TensorFlow Lite][14] 库，它与 MobileNets 和 Android 神经网络 API 是兼容的。
|
||||
|
||||
|
||||
[][15]
|
||||
**AIY 视觉套件包装图**
|
||||
（点击图像放大）
|
||||
|
||||
|
||||
除了提供这三个模型之外，AIY 视觉套件还提供了基本的 TensorFlow 代码和一个编译器，因此用户可以去开发自己的模型。另外，Python 开发者可以写一些新软件去定制 RGB 按钮颜色、压电元素声音、以及在 VisionBonnet 上的 4x GPIO 针脚，它可以添加另外的指示灯、按钮、或者伺服机构。Potential 模型包括识别食物、基于可视化输入来打开一个狗门、当你的汽车偏离车道时发出文本信息、或者根据识别到的人的面部表情来播放特定的音乐。
|
||||
|
||||
|
||||
[][16] [][17]
|
||||
**Myriad 2 VPU 结构图（左）和参考板**
|
||||
（点击图像放大）
|
||||
|
||||
|
||||
Movidius Myriad 2 处理器在一个标称 1W 的功耗下提供每秒万亿次浮点运算的性能。在被 Intel 收购之前，这个芯片最早出现在 Tango 项目的参考平台上，并内置在 2016 年 5 月由 Movidius 首次亮相的、Ubuntu 驱动的 USB 的 [Fathom][25] 神经网络处理棒中。根据 Movidius 的说法，Myriad 2 目前已经在 “市场上数百万的设备上使用”。
|
||||
|
||||
**更多信息**
|
||||
|
||||
AIY 视觉套件可以在 Micro Center 上预订，价格为 $44.99，预计在 12 月初发货。更多信息请参考 AIY 视觉套件的 [公告][18]、[Google 博客][19]、以及 [Micro Center 购物页面][20]。
|
||||
|
||||
--------------------------------------------------------------------------------
|
||||
|
||||
via: http://linuxgizmos.com/google-launches-tensorflow-based-vision-recognition-kit-for-rpi-zero-w/
|
||||
|
||||
作者：[ Eric Brown][a]
|
||||
译者：[qhwdw](https://github.com/qhwdw)
|
||||
校对：[校对者ID](https://github.com/校对者ID)
|
||||
|
||||
本文由 [LCTT](https://github.com/LCTT/TranslateProject) 原创编译，[Linux中国](https://linux.cn/) 荣誉推出
|
||||
|
||||
[a]:http://linuxgizmos.com/google-launches-tensorflow-based-vision-recognition-kit-for-rpi-zero-w/
|
||||
[1]:http://twitter.com/share?url=http://linuxgizmos.com/google-launches-tensorflow-based-vision-recognition-kit-for-rpi-zero-w/&text=Google%20launches%20TensorFlow-based%20vision%20recognition%20kit%20for%20RPi%20Zero%20W%20
|
||||
[2]:https://plus.google.com/share?url=http://linuxgizmos.com/google-launches-tensorflow-based-vision-recognition-kit-for-rpi-zero-w/
|
||||
[3]:http://www.facebook.com/sharer.php?u=http://linuxgizmos.com/google-launches-tensorflow-based-vision-recognition-kit-for-rpi-zero-w/
|
||||
[4]:http://www.linkedin.com/shareArticle?mini=true&url=http://linuxgizmos.com/google-launches-tensorflow-based-vision-recognition-kit-for-rpi-zero-w/
|
||||
[5]:http://reddit.com/submit?url=http://linuxgizmos.com/google-launches-tensorflow-based-vision-recognition-kit-for-rpi-zero-w/&title=Google%20launches%20TensorFlow-based%20vision%20recognition%20kit%20for%20RPi%20Zero%20W
|
||||
[6]:mailto:?subject=Google%20launches%20TensorFlow-based%20vision%20recognition%20kit%20for%20RPi%20Zero%20W&body=%20http://linuxgizmos.com/google-launches-tensorflow-based-vision-recognition-kit-for-rpi-zero-w/
|
||||
[7]:http://linuxgizmos.com/free-raspberry-pi-voice-kit-taps-google-assistant-sdk/
|
||||
[8]:http://linuxgizmos.com/google-releases-cloud-vision-api-with-demo-for-pi-based-robot/
|
||||
[9]:http://linuxgizmos.com/files/google_aiyvisionkit.jpg
|
||||
[10]:http://linuxgizmos.com/files/rpi_zerow.jpg
|
||||
[11]:http://linuxgizmos.com/raspberry-pi-cameras-jump-to-8mp-keep-25-dollar-price/
|
||||
[12]:http://linuxgizmos.com/files/google_aiyvisionkit_pieces.jpg
|
||||
[13]:http://linuxgizmos.com/files/google_visionbonnet.jpg
|
||||
[14]:https://developers.googleblog.com/2017/11/announcing-tensorflow-lite.html
|
||||
[15]:http://linuxgizmos.com/files/google_aiyvisionkit_assembly.jpg
|
||||
[16]:http://linuxgizmos.com/files/movidius_myriad2vpu_block.jpg
|
||||
[17]:http://linuxgizmos.com/files/movidius_myriad2_reference_board.jpg
|
||||
[18]:https://blog.google/topics/machine-learning/introducing-aiy-vision-kit-make-devices-see/
|
||||
[19]:https://developers.googleblog.com/2017/11/introducing-aiy-vision-kit-add-computer.html
|
||||
[20]:http://www.microcenter.com/site/content/Google_AIY.aspx?ekw=aiy&rd=1
|
||||
[21]:http://linuxgizmos.com/raspberry-pi-zero-w-adds-wifi-and-bluetooth-for-only-5-more/
|
||||
[22]:https://www.movidius.com/solutions/vision-processing-unit
|
||||
[23]:https://www.tensorflow.org/
|
||||
[24]:https://research.googleblog.com/2017/06/mobilenets-open-source-models-for.html
|
||||
[25]:http://linuxgizmos.com/usb-stick-brings-neural-computing-functions-to-devices/
|
||||
[26]:http://linuxgizmos.com/google-launches-tensorflow-based-vision-recognition-kit-for-rpi-zero-w/
|
||||
Loading…
Reference in New Issue
Block a user