Sensor Light,Sensor Downlight,Sensor Ceiling Light,Sensor Garage Lamp JIANGMEN MOSCOT OPTOELECTRONIC TECHNOLOGY CO.,LTD. , https://www.sensorsled.com
Six New Features of Google's New Image Recognition Tool Break Through
On Wednesday, Google announced that they have developed a powerful image recognition tool called Google Cloud VisionAPI, which may be a very useful tool for developers. With it, developers can make their own software. The robot knows the information on the image.
Powerful image detection tool
First of all, as a machine learning domain technology, this tool is equipped with the intelligent machine learning system Tensor Flow that we introduced to you before. Google just opened the system to developers for free last month. Second, its main function is to learn and identify the content in the picture. The operating principle of this tool is not complicated. It can quickly divide pictures into thousands of large categories, such as: boats, lions, the Eiffel Tower, etc.; and then detect facial related emotions and identify related pictures. Text. In their opinion, the Google Cloud Vision API has the following attractive features:
Detection entity: The Google Cloud Vision API can detect entities in a picture from thousands of image categories. In addition, more attractive is that you can also create your own metadata based on your own image categories, so that you do not have to be subject to the thousands of categories that Google has set for you. Your own picture classification.
Character recognition: The optical character recognition technology of Google Cloud Vision API can recognize the text content on the picture. There are many supported languages.
Security Detection: Google Cloud Vision API security detection feature can identify inappropriate content in the picture, this feature is installed in Google Safe Search. It helps users avoid uncomfortable content.
Face Detection: The Google Cloud Vision API's facial recognition technology can detect portraits in images, as well as identify specific organs such as eyes and mouth, and more importantly, it can also detect happiness, sadness, etc. The 8 emotions. In order to protect the privacy of users, Google also promised not to store any information about the face in any Google server.
Landmark detection: The Google Cloud Vision API's landmark detection feature recognizes some of the world's most famous natural and cultural buildings and landscapes. It also shows you the latitude and longitude of the site.
Logo detection: Google Cloud Vision API's logo monitoring feature can detect the famous brand logo in the picture.
Here are some examples:
Google's breakthrough <br> <br> With this tool, the ability to machine recognition of images will be greatly enhanced. Before, you couldn't let the computer automatically filter the photos for you; but now, just enter the appropriate "category" and the computer can pick out the corresponding picture for you.
Developers can now design their own software and robots. In the near future, perhaps when we enter the house, the robot in the home can select the greeting it says based on the expression on your face, and you can search for the desired picture more accurately when you are on the Internet.
Of course, it is normal for Google to launch this tool. In this area, both Amazon and Microsoft have gone very far. Many of my friends may remember Microsoft’s previously introduced gadgets that identify age and mood, all from Microsoft’s Oxford program that focuses on artificial intelligence. Of course, Google cannot fall behind. In the past, Google has also launched some API products for prediction, but all of them have limited functionality and do not involve images. Therefore, this move is also a big breakthrough for Google.
This API currently offers some limited previews on the Google Cloud Platform, but in the future Google will add this feature to Google Cloud Storage. This can make it more widely used.
Science and technology dynamic image recognition is considered as one of the core technologies of several key industries in the future technology field. Microsoft, Google, Facebook, Amazon, Baidu, Tencent and other giants are pouring a lot of resources to promote this feature. Recent breakthroughs have confirmed that computers have the full potential to catch up to and even surpass humans in the field of image recognition, and also to enrich the confidence of practitioners. Over time, we will find that the computing devices around us are becoming more and more "smart."