Learn about the advantages of Unity machine learning technology combined with the gaming industry, applications in other areas, and future developments.
Unity has been working to lower the threshold for machine learning and apply it to game development to make the development process easier. Unity ML-Agents, a Unity machine learning-based product, is an open source AI toolkit that helps developers and researchers train machine agents in real and complex environments.
At the recent Unite Beijing 2018 conference, Unity AI and Danny Lange, vice president of machine learning, shared Unity's new developments and related applications in machine learning. After the sharing, we interviewed Danny Lange to gain an in-depth understanding of the combination of Unity machine learning technology and the gaming industry, applications in other fields, and future developments.
The advantages of combining machine learning with the gaming industry
Danny Lange has a lot of confidence in machine learning over the years. He believes that machine learning can bring subversive changes to human life. Previously Danny Lange tried to integrate machine learning in different fields, and now he has entered the game industry and joined Unity.
"Our current practice is to make these algorithms, which are difficult for ordinary people to accept, to be an algorithm that all developers can use," Danny Lange said.
In other words, Danny hopes to lower the threshold for machine learning, and based on Unity's huge developer base, allows more game developers to use AI to develop games.
Danny Lange said that one of the advantages of machine learning applications in game development is to help developers create NPCs. We all know that the language and behavior of NPC in the game are more modular. If you want to create a smart looking NPC, you need better algorithms to support, and a better algorithm means higher costs.
“Machine learning can create NPCs with multiple behaviors. After repeated training through machine learning, PCs can learn many human behaviors, which makes machine-human interactions more natural,†says Danny Lange. Using machine learning to create NPCs can make NPCs smarter, reducing the cost of game development.
According to Danny Lange, in addition to creating NPCs, in a completely different field, that is, applying machine learning to pre-release testing, Unity's machine learning product Unity ML-Agent can be used instead of humans to test games. Help developers understand if the game is going smoothly, and increase efficiency while reducing the cost of testing.
Specific application of machine learning
Based on Unity, machine learning can obviously bring a lot of benefits to the game, the most obvious is to improve gameplay. Danny Lange said: "The core of game AI is to bring fun. Therefore, AI can learn to optimize the fun of the player. For the player, there will be more personalized and customized things in the game. â€
“As we saw in Reinforcement Learning, if you operate correctly, ML-Agent can learn more strategically. That is, it does not kill the player immediately, but rather seduce the enemy. The player will feel that he can get rid of it right away, but in the end, the player feels that the game is very fun."
Currently, Unity's machine learning technology is mainly used in the game field, but according to Danny Lange, there are many areas where machine learning technology can be applied.
“For machine learning, autonomous driving is a promising area. After applying machine learning technology, people no longer need to drive on the real road as they actually drive, because that could lead to problems. People Just simulate in the virtual space."
“In another area, if you are a student or a researcher, you can apply the algorithm of machine learning to research and teaching. Currently, in American universities, our machine learning tools are very popular. They use our tool as a case study for machine learning in universities."
“From an architectural point of view, you can also apply machine learning. For example, when designing a channel, using machine learning to rationalize the design of the channel traffic. For example, this building may be confusing, but in When designing, use machine learning to simulate, where people go from, where to go, what path to choose, etc. Through machine learning, the design of the path is more convenient and reasonable."
Unity's future development in machine learning
If the historical significance of the industrial era is to liberate the productive forces, then our current era should be regarded as the era of artificial intelligence, using computer AI technology to assist us in accomplishing more difficult things that human beings themselves can accomplish. From Siri to AlphaGo, smart home, smart healthcare, and autonomous driving, artificial intelligence has begun to affect every aspect of our social life.
Machine learning is an important branch of discipline generated from artificial intelligence and is the key to achieving intelligence. Grasping the development of machine learning, we can follow the wave of artificial intelligence development, so how to make better use of machine learning in the future is very important.
Speaking of this, Danny Lange said: "In fact, in the United States, companies like Amazon, Google, and Apple have used machine learning to replace traditional software and code. Unity hopes that through our platform, we can Machine learning is introduced to more developers and they are allowed to use machine learning."
CCTV Power Supply,CCTV Power Supply Box,CCTV DC Distributed Power Box
Chinasky Electronics Co., Ltd. , https://www.cctv-products.com