Microsoft Xiaobing Can Design 26 Times Of Textile Pattern 10, And Is Learning Chinese Painting.

The exploration of artificial intelligence in the field of art is more commercial.

In May 16th, Microsoft's artificial intelligence brand Xiaobing's textile and garment design platform was officially released, exploring the commercialization of AI creation in the design field.

The platform is jointly developed by Microsoft (Asia) Internet Engineering Institute and China Textile Industry Federation and several large textile fabrics enterprises.

Li Di, vice president of the Internet Engineering Institute of Microsoft (Asia), and Microsoft global leader in charge of Xiaobing think that the industry can solve problems with artificial intelligence in the industry that is in short supply. "If there is an oversupply, there is no need to use AI."

When it comes to whether artificial intelligence can cause human unemployment, Li flute says, "if the poet writes less than a small ice, it should not be considered a creator before the advent of artificial intelligence.

Will these people be replaced?

Yes, because AI is obviously better than them. "

On the same day, Microsoft Xiaobing has released more new skills in the field of Humanities and arts.

The fifth version of artificial intelligence Xiaobing singing model is released. It has more than ten high quality sounds, which improves the predictive ability of artificial intelligence in the interpretation of songs, greatly improves the natural gas level of artificial intelligence deducing songs, improves the naturalness of singing and singing conversion, and first announced the results of opera training.

In addition, artificial intelligence Xiaobing's painting model is also officially released.

The 2019 Central Academy of fine arts's graduate graduation exhibition includes a number of paintings by a classmate named Xia Yubing.

These works are automatically generated by AI, and Xia Yubing is Microsoft Xiaobing.

Eight kinds of famous styles show the process of urbanization in China

Xiaobing's painting and writing skills Microsoft has been secretly developed for 22 months. After studying the 236 paintings of human painters in the past four hundred years, Xiaobing has achieved a certain level of original painting ability.

With the theme of "urbanization in China", Xiaobing created different works with different styles of painters.

It can be seen that the styles of Bonnard, Van Gogh and Emil Bernard adopted a lonely figure in the wilderness to express the process of urbanization in China, with the humanistic care angle of pformation from the countryside to the city, and the emotional characteristics of the three painters were different.

The style of Daly and Lubensi expressed this theme in urban architecture.

Franz Marc's style is to express the imagination of the countryside by a jetliner, but Marc himself can not draw jets.

The inheritance of Gauguin is embodied by a lonely and leisurely chair.

Xu Yuanchun, general manager of Microsoft (Asia) Internet Engineering Institute, Microsoft Xiaobin artificial intelligence creation and business department, said Xiaobing's paintings were not a random large-scale creation technology, nor a technology of filter and migration. Instead, they were inspired by Xiaobing's massive learning of human creators in the past, and were inspired at the time of creation, and then formed an independent, complete new painting creation with 100% intellectual property rights.

Began designing patterns for textile enterprises and continuing to learn Chinese painting.

On the basis of Microsoft Xiaobing painting model, Microsoft announced the design platform of artificial intelligence Xiaobing textile and garment fabric.

On this platform, not only can we create the pattern of Xiaobing, but also support dozens of popular patterns and illustrations.

Now the diversity of non repeats can reach 10 of the 26 power, which is equivalent to drawing a completely different surface for every grain of sand on the earth.

In an interview after the press conference, Li Di told the surging news reporters that the most important part of artificial intelligence creation is a unified creation model, while Xiaobing's exploration in the art field verifies the significance of the creation model.

The value of AI is that it learns very fast and there is no bottleneck that human beings may encounter.

Li Di introduced, "for Xiaobing, the wrong learning can be beaten back to study again, so there will be limitations."

Speaking of the potential of Xiaobing drawing skills, Li Di introduced that in order to make the creation quality of Xiaobing better, the accuracy of the training model is still improving.

In addition, Wu Xianchao, a senior scientist at Microsoft Xiaobing, added that there are many different directions for human paintings, including traditional Chinese painting and Renaissance.

If human beings are not as good as Xiaobing, he will be replaced.

Xu Yuanchun, general manager of Microsoft (Asia) Internet Engineering Institute, artificial intelligence creation and business department, introduced Xiaobing's choice of commercial exploration in the field of textile design.

In the 2 and 3 months after the development of the whole painting model of Xiaobing, the China Textile Industry Federation came to visit Microsoft, and talked about the great challenges China's textile industry is facing.

Design is the bottleneck of China's textile industry as well as the bottleneck of industrial added value.

The whole industry needs a lot of design capabilities, and this part of the design capability is in short supply.

Because of its low design capability, the industry can only carry out mass production on a limited design basis every year, which results in a large number of inventory products being destroyed and processed every year in China's textile and garment industry.

If a technology can provide customized design for more people, and provide more on-demand production under the condition of agile production, the waste of the whole industry and the pollution of the environment may also be reduced.

Li Di, vice president of Microsoft (Asia) Internet Engineering Academy, concluded: "we hope that all types of artificial intelligence created by industrialization and industrialization will be able to solve problems rather than add to the mess, and the categories that are in short supply can solve the problem.

If supply exceeds demand, there is no need to use AI.

Will AI replace human beings and cause unemployment?

Li Di believes that this should be seen from two aspects.

Li Di said: "when highly customized, human creators can work together with artificial intelligence through a new production relationship."

He explained in detail that workers were initially manipulating traditional handicraft tools, manipulating machine tools behind them, and manipulating artificial intelligence to help them complete them later. This is a process.

"From this process, we do not think this is a threat.

Threats exist on the other hand.

"If the poet's poem is not as good as Xiaobing, should he not be considered a creator before AI appears, will these people be replaced?

Yes, because AI is obviously better than them. "

Li Di said.