您现在的位置是:DeepMind introduces self >>正文
DeepMind introduces self
上海品茶网 - 夜上海最新论坛社区 - 上海千花论坛9人已围观
简介By subscribing, you agree to our Terms of Use and Policies You may unsubscribe at any time.Robots to...
By subscribing, you agree to our Terms of Use and Policies You may unsubscribe at any time.
Robots today can do a variety of tasks as long as they are trained on real-world data. But what if they could bypass this step? It would result in many more general-purpose robots being developed at a faster rate.
Google’s DeepMind has introduced a self-improving AI model called RoboCat that may just be the key to machines that can self-generate new training data to improve their technique without too much human interference.
This is according to a blog by the firm published on Tuesday.
“RoboCat learns much faster than other state-of-the-art models. It can pick up a new task with as few as 100 demonstrations because it draws from a large and diverse dataset. This capability will help accelerate robotics research, as it reduces the need for human-supervised training, and is an important step towards creating a general-purpose robot,” state the researchers in their post.
See Also Related- Google forms AI dream team by merging 'Brain' and 'DeepMind' projects
- Google's DeepMind: Optimized algorithms not trained on human code
- Google and DeepMind just launched MedPaLM, a large language model
They go on to explain that the new model is based on DeepMind’s multimodal model Gato (Spanish for “cat” which explains the name). This model can process language, images, and actions in both simulated and physical environments.
The researchers made use of Gato’s architecture that comes with a large training dataset of sequences of images and actions of various robot arms solving hundreds of different tasks.
They then trained RobotCat to learn new tasks by following five steps:
1. Collect 100-1000 demonstrations of a new task or robot, using a robotic arm controlled by a human.
2. Fine-tune RoboCat on this new task/arm, creating a specialized spin-off agent.
3. The spin-off agent practices on this new task/arm an average of 10,000 times, generating more training data.
4. Incorporate the demonstration data and self-generated data into RoboCat’s existing training dataset.
5. Train a new version of RoboCat on the new training dataset.
This diverse training taught the AI model to operate different robotic arms within a few hours. And RobotCat was quick to adapt. Even though it had not been trained on arms with two-pronged grippers, it was able to adapt to a more complex arm with a three-fingered gripper and twice as many controllable inputs.
The more new tasks it learned, the better it got at learning additional new tasks. Early versions of RoboCat were successful just 36 percent of the time on previously unseen tasks. However, the latest and most advanced RoboCat, which had trained on a greater diversity of tasks, more than doubled this success rate on the same tasks.
Tags:
转载:欢迎各位朋友分享到网络,但转载请说明文章出处“上海品茶网 - 夜上海最新论坛社区 - 上海千花论坛”。http://www.jz08.com.cn/news/4849.html
相关文章
DARPA project 'Blackjack' downsized to four satellites only
DeepMind introduces selfBy subscribing, you agree to our Terms of Use and Policies You may unsubscribe at any time.On Monday...
阅读更多
The Top 3 Alternatives to Ethereum that you can buy on December 22: DOT, ZEC, and LINK.
DeepMind introduces selfPolkadot (DOT) saw an increase in value of 7.40% in the last 24 hours.Zcash (ZEC) saw an increase in...
阅读更多
Marathon Digital buys another 30,000 Bitcoin miners
DeepMind introduces selfThe North American Bitcoin mining company has bought Antminer S19j Pros from the Beijing-based compa...
阅读更多
热门文章
- Fast 6G wireless communication achieved, new study shows
- This new simulation platform trains robots to hear their environment
- Neural Lab’s air gesture technology could be a game
- Dream Chaser spaceplane aces key test, could soon fly to orbit
- MIT's new liquid metal printing can build chairs in minutes
- Where to buy COCOS: trending today, gained 34% in 24 hours
最新文章
Enter the world of Internet
Three best cryptocurrencies to buy with less than $1
You can now buy GALA, the native token of Gala Games: here’s where
IMF Managing Director backs the US dollar over Bitcoin
Looking for exciting blockchain opportunities: The best DApps to Invest in today
Boeing's Starliner astronaut launch delayed indefinitely due to two major issues