Japan still has significant highlights in the application of AI technology, or that the brain is strange.
Editor’s note: This article is from the WeChat public account “Gate Ventures” (ID: Thejiangmen), author Ace.
A few days ago, the teenager watched a video. A research team from Keio University in Japan designed a robot that would help the elderly who are walking unstable to maintain balance, and this robot is a “tail.”
This one-meter machine is called Arque. The reason why it is designed as a tail is to use bionics. For example, the cheetah’s tail can balance it while running and climbing. The Arque installed on the person “can swing like a pendulum. When the human body leans on the body, the tail moves in the opposite direction.”
Does it look familiar?
This is simply a big monster wow! Since then, the curiosity of the teenager has been like a raging fire. I decided to see what wonderful AI applications have been invented in Japan?
To remind everyone here, Japan gives a lot of people a high-tech impression, but in terms of AIJapan’s research in the field of artificial intelligence is relatively backward between China and the United States. Sun Zhengyi, chairman of Japan’s Softbank, once said that Japan has become a “developing country” in the artificial intelligence revolution.
Despite this, the teenager found that Japan’s application level of AI technology is still bright, or that the brain is strange. After a massive search and reading, the teenager carefully selected and summed up a mini “Japanese AI singular record”, are you ready to see the official?
Japan AI Reports
异闻一:You are willful! Million dollars to create “AI Buddha”
The Japanese have an obsessive fanaticism about Buddhism. After all, more than 80% are Buddhists. This belief also spread to the AI world. In order to stimulate the interest of young people and foreigners in Japanese Buddhism, the 400-year-old Gaotai Temple cooperated with Osaka University to develop an “AI Buddha” Mindar.
This 6-foot-tall robot is worth nearly $1 million and is made of silicon and aluminum. It is built in the image of Kannon (the god of mercy in Japan) and can recite the “heart”. It is also the first robotic god in the world to combine spiritual and scientific fields. Recently, Mindar has settled in Gaotai Temple and plans to conduct Buddhist sermons.
This scene can’t help but think of the movie “The Report of Human Death”. There is a story “Tiangong Kaiwu”: the temple introduced a robot, which was originally produced for cleaning work, and passed through the temple. After a while, the drums of the morning bell suddenly had their own consciousness. “Enlightenment into a Buddha” made the human beings who relied too much on robots panic…
The “Destruction of Humanity – Tiangong Kaiwu” Stills
异闻二:The study of Buddhism in Nara University, Japan: Find your true Buddha
If there is no money and can’t be willful, there is a Buddha expression recognizer here. A group of students at Nara University in Japan recently developed an artificial intelligence project to use AI to decode Buddha images to judge the mood and age of different Buddha images. The best part is to upload your own photos to find the one that best matches your mood.
Be sure to go to this website, full of Buddhist technology, and try the big family, the effect is as follows
The deity is like a Buddha
Father and Son
Come on the door sisters~
Trying to try out the new Buddha image several times
Let us congratulate this friend for brushing an unusual Buddha image from a singular perspective
In order not to distract everyone’s attention, the teenager put the link and usage guide to the end
Being able to make such a Buddhist study can be inseparable from Japanese Buddhist culture. Believers believe that the expression of a Buddha image is entirely determined by the individual’s mood. Traditionally, Buddha statues should not have labels such as age, emotion, and gender. When making Buddha images, the craftsmen were also told that they could not give the Buddha any human facial expressions. Therefore, based on beliefs and other factors, the best interpretation of the Buddha’s emotional content is not human.
Who can it be?
This research group used Microsoft’s AI system Cognitive Services to quantify the eight expressions of anger, horror, joy, neutrality, and sadness through the Face API to analyze its tendency. They conducted an emotional analysis of 216 Buddha images throughout Japan and published survey results of the estimated age and expression of the Buddha image. The results show that the average age of the Buddha statue is 38 years old. The youngest is the statue of Sakyamuni in Zhengfa Temple (2.9 years old), and the statue of Rui Zun in Xi Da Temple is the throne of the oldest Buddha statue in 83.5 years old.
Sekine, a professor who led 18 students to the study, said, “If you let the world judge the Buddha is a kind ofOh, then it might be fair to entrust this process to artificial intelligence.
异闻三;Deep learning development “Boss is coming” alarm
The term “social animals” originated in Japan and was the self-deprecating language of the bottom employees of the company. It first appeared in Japan in the 1990s and then became popular in the East Asian world.
There are always some bosses who have thieves in their eyes. It seems that they have walked around with carelessness. In fact, Yu Guang has already seen your screen. When you go to work, you want to fish in troubled waters (such as you are reading this article at this time, or you are planning to pull to the end of the text to see which Buddha you are). You must always pay attention to the boss’s dynamics. Some skilled people can quickly switch screens, but the fastness of hands does not mean that the expressions are not shocked. Sometimes, the troubles may create the effect of “no silver three hundred two here”.
What should I do? A Japanese programmer decided to write a program once and for all. The idea is as follows: First, let the computer complete the deep learning of the boss’s face, use the camera to automatically identify the face that has passed through the station channel. If it is confirmed to be the boss, use a screenshot full of code to cover the entire screen. To show that I am working hard.
The process is not complicated and is divided into three steps.
In order to save space, the teenager made a map
Following friends can download the source code here:
https://github.com/Hironsan/BossSensor
异闻四:Change the light with a ringing finger, the machine learning assisting sister of the technical house
Not all programmers are full of work and ideals, but actually sisters.
This program is simpler than the boss recognizer, with only two steps: 1. Identifying the sound of the ring finger; 2. Changing the light bulb. According to the classmate’s technical blog, his logic is to detect the sound of the finger, and naturally he wants to detect the single pronunciation (the sound that suddenly becomes louder in an instant), but the problem that is easy to occur is that the ringing sound can be compared with other The single-sounding pulse grows almost the same. How do you let the program learn to turn on the light only when you are firing the finger?
Small brother used spectrum analysis to extract the difference between the single vocalization of the slap and the knocking table. The following figure can find more significant differences after spectrum analysis:
Not only that, he broke the heart of the single dog, the end of the article introduced the actual application of the program (捂 face.jpg)
Please use this product when you take the girl home and enter the ambiguous atmosphere.
1, just when the two people talked about nothing to talk about, the atmosphere suddenly entered a state of paralysis, hit a ring finger
2, single-sound spectrum analysis to extract sound features
3. Machine learning recognizes and judges the ring finger
4, then the light bulb changed from white to pink and sent to assist
5, after that… give it to yourself to play
He also frankly: Actually, this is a thing that doesn’t make much sense. But doing something that seems to have no meaning at all, enjoying the production process is also one of the joys of siege lions.
异闻五:Seeing your thief’s eyebrows, it’s not a good thing at a glance
A Japanese startup called Vaak claims to have developed an artificial intelligence software that analyzes the surveillance videos in the store and discovers suspicious behaviors suspected of stealing, identifying them before the thief “hands-on” and sending alerts to store staff. In order to achieve the purpose of crime prevention.
VaakEye, a smart stealing warning system, analyzed the thief’s various modes of action after more than 100,000 hours of video training.