In the progress of constructing an intelligent future, datasets play a vital role. From autonomous driving cars to smart security systems, high-quality datasets provide AI models with massive amount of learning materiel, empowering AI model more adaptable in various real-world scenarios. Companies and researchers through continuously improving the efficiency of data collection and annotation can accelerate the implementation of AI technology, help all industries achieve their digital transformation.

Art is obviously not just a simple calculation problem. Different artists have different landscapes and different ways of expression. And AI is just a cold machine, everything is just a bunch of data in his eyes, so people firmly believe that AI can’t create good works of art.

But the reality is quite the opposite. AI creates art by generating adversarial network models. This model consists of two competing parts. One is the generator, which keeps creating, and the other is the discriminator, which is to analyze and determine which is the “real” works from the data set and which is “false” works from the machine. When the generated works successfully “cheats” the discriminator, the model training is done.

The popular chatbot Xiaoice, created by Microsoft, can accomplish original paintings independently, after she digested the paintings of 236 famous painters in the past four centuries of art history. The originality of her paintings is not only reflected in the composition, but also in the use of color, expressiveness, and the detailed elements contained in the works.

According to Li Di, the vice president of Microsoft’s Software and Technology Center of Asia, only 30% of Xiaoice’s works can be traced to the source of whom she learned from, and the rest cannot be traced completely. “You don’t know what she is thinking. AI is like this. When you decide to use a deep neural network, you have to accept that there is a part of her that is a mistery”, Li Di said.

Besides painting, Xiaoice has also made great achievements in the creation of other art forms. In May 2017, Xiaoice published the first collection of poems “Sunshine Misses Windows” created by AI. In 2019, Xiaoice and human authors jointly created the poem “Flowers are the Silence of Green Water”, which is also the world’s first literary work jointly created by AI and humans.

It is worth mentioning that on June 29, 2020, Xiaoice graduated with her human classmates as assessed by the Music Engineering Department of the Shanghai Conservatory of Music, and was awarded the “Honorary Graduate” of the 2020.

As a new method of art creation, AI creation is different from the general single imitation of human’s intelligence, but presents a vigorous situation where human-machine collaboration continues to deepen and the quality of works continues to improve. The creative practice of AI has also objectively promoted the changes in the existing art production methods, and has made the necessary technical and practical pavement for the new art form.

On-demand Data Services

Nexdata has professional data collection equipment, tools and environment. Our project managers have extensive experiences in data collection and quality control. We can meet various scenarios and types of data collection needs. As for data annotation, Nexdata has 3 mega-data bases and more than 5,000 professional annotators, supporting on-demand data annotation services, such as speech, image, video, point cloud and text, etc.

Technological progress will bring subversive changes to the art world, and it’s worth thinking about how aesthetic art can be reborn in the new round of technological revolution is worth. In the future, human and intelligent machine will achieve a high degree of integration, and the application effect of AI technology in art will exceed the current human cognition of art.

End

If you need data services, please feel free to contact us: info@nexdata.ai

With the rapid development of artificial intelligence, the importance of datasets has become prominent. By accurate data annotation and scientific data collection, we can improve the performance of AI model, which enable them to cope with real application challenges. In the future, all fields of data-driven innovation will continue to drive intelligence and achieve business results in high-value.