Artificial intelligence generally appears as a tool to assist human work, sweeping robots, medical robots, waiter robots, etc. are the most common forms of artificial intelligence. In fact, artificial intelligence is not the only form of robot, from the field, including robotics, language recognition, image recognition, natural language processing and expert systems. Today we will look at, in addition to robots, what are the other products of artificial intelligence.
Google's artificial intelligence project DeepMind
Google's London-based research and development arm DeepMind has developed AI technology that can play video games autonomously. Computer systems based on DeepMind's technology are able to learn at an astonishing rate, quickly mastering gameplay and becoming proficient in game-winning methods. Previously, the team called the depth of Q-network learning network, only need to observe the game screen and the game score changes, you can analyze the "pass skills" and get high scores of the gameplay and algorithms, can reach the level of professional human players.
Google's artificial intelligence project DeepMind
The system has been tested on 49 games with the same algorithms, network architecture, and parameters, and is now proficient at 22 games (including the aforementioned SpaceInvaders), reaching expert levels of play. This system further proves that AI can learn through deep learning so as to master gaming skills and gain the same control as humans, even surpassing them in some aspects.
IBMWatson
Last year, IBM released WatsonAnalytics. watsonAnalytics implements a natural language-based cognitive service that delivers instant predictive and visual analytics tools to businesspeople. watsonAnalytics will be available as a free, cloud-based service by the end of this year. value-added application version (FreemiumVersion) available on computers and mobile devices by the end of this year.
IBMWatson
WatsonAnalytics provides self-service analytics capabilities, including data access, data cleansing, and data warehousing, to help business users acquire and prepare data, analyze it, visualize the results, and provide a foundation for users to take effective action and further interaction.
Microsoft AI Torque Chinese version
In February this year, Microsoft released the Chinese version of Torque, a gesture-driven, voice-interactive AI product tailored for Chinese users on the Android platform. As Microsoft's first AI product on the Android platform, and also Microsoft's first Chinese product for wearable devices, Torque's goal is to use the smallest interface to deliver information in the most direct and timely manner, and Torque's birth frees up the Android user's hands, so that the user only needs to gently shake his wrist, and then say, "Happy Camp! The user only needs to flick his wrist and say to it: "Happy camp host", "Where is the nearest KFC", "Call Zhang Yong" and other commands, and then he can experience getting information and accomplishing more tasks in a minimalist way - this is precisely Microsoft's vision of mobile Internet. -This is Microsoft's understanding of mobile productivity and efficiency in the age of mobile connectivity.
Microsoft Torque Chinese version
According to Microsoft, Torque, like Xiaobing, Xiaona and other Microsoft AI products, uses the Bing Big Data platform as the underlying engine to process the voice commands that each user uploads to the cloud through their cell phones and the mobile Internet; and the AI product team at Microsoft's Internet Engineering Institute (Asia) has developed an AI product for Chinese users' preferences and habits. The AI product team from Microsoft's (Asia) Internet Engineering Institute has made special design and localized development on the functions for Chinese users' preferences and habits.
Youtube Automatic Captions
Google has been using its voice recognition technology to provide real-time "Automatic Captions" on YouTube since 2009, allowing users to watch thousands of videos on the web without turning on the speakers and without disturbing others.
Youtube Automatic Captions
YouTube adds captions to YouTube videos using Google's Automatic Speech Recognition (ASR) technology, which is derived from GoogleVoice, and while the resulting captions are unlikely to be 100 percent accurate, they can at least help you improve your listening skills so you can understand the content of the video. Google is always improving its automatic speech recognition technology. This technology supports automatic subtitles in English, Japanese, Korean, Spanish, German, Italian, French, Portuguese, Russian, and Dutch.
In addition to the automatic subtitle feature, YouTube has added subtitle time and auto time for those who subtitle their videos, making it easier for people to make their own subtitles. All you need to do is create a simple text file with all the words spoken in the video, and then Google, using automatic speech recognition technology, can match those words in the text to the words it recognizes itself, so that accuracy is improved and you don't have to spend too much time subtitling sentence by sentence.
Artificially Intelligent Bionic Eye
The world's first artificial bionic robotic eye transplant for the treatment of blindness caused by age-related retinal macular degeneration (AMD) has been successfully carried out at the Royal Eye Hospital in Manchester, UK. The artificially intelligent bionic eye device, known as ArgusII, consists of two parts: an in vivo implantable part and an out-of-vivo part that the patient must wear. The implanted device, which will be implanted into the patient's retina, contains an array of electrodes, batteries and a wireless antenna. The external device contains a pair of eyeglasses with a built-in forward-facing camera and radio transmitter and a video processing unit.
Artificially Intelligent Bionic Eye
The camera captures the image directly across from the implant, sending the signal to the video processor to await processing. The processed signal is sent back to the glasses, where it is picked up by the antenna of the implanted device. Ultimately, the video is "output" to the electrode array, which acts as an optic nerve simulator. The resolution of the electrode array is at the 60-pixel level, which is more than enough to allow the implanted device to track the movement of an object, see basic patterns and shapes, or slowly read larger text. the image provided by ArgusII is in black-and-white, but the Argus development team is working on compiling the brain stimulation from the electrodes in the hope of having the brain receive color signals as soon as possible.
Patients recovering from the surgery have been able to recognize vertical or horizontal lines, recognize faces, and read newspapers without magnification. What's more interesting is that with this surgery, patients are able to see the view of their eyes even when they are closed, which makes it somewhat interesting.
Additionally, an AI eyeball developed by Second Sight, a US-based company that develops AI eyeballs, has been approved for marketing, which can restore sight to blind people who are completely blind.
News-writing robots
News-writing robots
The Associated Press has been using the Wordsmith platform since last summer to automate financial news writing. According to Lou Ferrara, AP's director of business news, the use of algorithm-based machine news writing will increase the amount of quarterly business news stories about companies in AP's business news by more than 10-fold, from 300 to 4,400 stories per quarter, without adding new staff, while at the same time "freeing up" journalists previously used for such stories. At the same time, it will "free up" the journalists who were previously doing this type of reporting to do more creative and challenging work in planning and sourcing news. When the system first went live, it was still necessary to manually review articles and make adjustments to the platform; three months later, it has completely eliminated the need for human intervention.
Bird Face Recognition Technology Developed at Cornell University
Cornell University, in conjunction with the VIsipedia Research Initiative group***, has developed the MerlinBirdPhotoID software, which can be used to identify the species of birds that appear in various images with the help of computer vision recognition technology and deep machine learning. This is great software for introductory bird watchers and bird enthusiasts. With deep machine learning, this program is able to provide recognition results within seconds, and the accuracy of the first three recognition results has reached more than 90 percent.
Cornell University develops bird face recognition technology
Users can narrow down the identification by uploading pictures of birds whose species are unknown and by boxing out the images of the birds that need to be recognized. The software is able to point out known species of birds from tens of thousands of images, and the database currently contains more than 400 species of birds common in North America. With the number of users, and deep machine learning, the accuracy will increase day by day. "Computers can process images much more efficiently than humans, and they are able to categorize, index, and process a large number of graphically detailed features to identify results," said Serge Belongie, a professor at Cornell University.
SkypeTranslator
Microsoft's real-time translation tool, SkypeTranslator, combines speech-recognition technology with what Microsoft calls a "deep neural network and Microsoft's proven static machine-translation technology". It automatically translates voice calls and instant messaging messages in different languages.
Currently, English, Spanish, Italian and Mandarin Chinese are supported. In addition, 50 languages are supported for instant messaging, including French, Japanese, Arabic, Welsh, and even Klingon.
SkypeTranslator
This translation tool is widely favored because it integrates advanced technologies such as machine translation, speech recognition, machine learning and big data. It is understood that the speech recognition technology required for real-time Skype Chinese interpretation is jointly developed by researchers from Microsoft China and the United States.