Meta is on the lookout for innovation and its recently talked about ‘Make-a-Video’ AI system is all set to revolutionize the world of video content.
The breakthrough technology was first talked about in July of this year and it was compared to the likes of Dall-E. Here, machine-based learning is used to bring ahead algorithms to make great depictions of written material.
Then on Thursday, the firm’s CEO spoke about how it’s going to be releasing an animated contemporary of its already popular ‘Make-A-Scene’.
As you may probably guess by the name, this is a new system that allows users to turn text into short-form video clips of the best quality as Zuckerberg bragged about in his blog on Thursday.
And the functioning is very similar to how Scene works, mixing different language processing and networks to use non-visual prompts to form pictures. In the end, it’s just like putting content into a new and unique format.
Meta says the idea behind the technology is very clear-cut and simple. It’s all about learning what the world appears like and how its best envisioned through the pairing of text with images. You also learn more about how it moves via unsupervised videos. This news came from the company’s team of researchers who published a paper on the findings of the new project.
When used correctly, the amount of time needed to have your video models trained is dramatically reduced. And it also gets rid of the need for paired text with video data sources too. All in all, you’ve got the diversity and depictions mastered brilliantly that can be found in the best generation models of today.
Just like similar projects coming forward from Meta’s AI team of researchers, this one is also going to be open-sourced. The company adds that it’s taking careful steps to build such innovative systems and share the results of the research with the world to harness feedback. They also vowed to better finetune and transform the approach to raging technology.
Just like any other AI release that comes with time, this opportunity for it to be misused can arise. And for this particular launch, it might not be a huge one. Hence, to rid any chances of malicious shenanigans, the team has eliminated any form of toxic phrasing and training dataset that comes along with it.
Read next: Meta Will Make History As It Plans Hiring Freeze, Reduced Headcount, And Team Reorganization For Facebook
The breakthrough technology was first talked about in July of this year and it was compared to the likes of Dall-E. Here, machine-based learning is used to bring ahead algorithms to make great depictions of written material.
Then on Thursday, the firm’s CEO spoke about how it’s going to be releasing an animated contemporary of its already popular ‘Make-A-Scene’.
We’re pleased to introduce Make-A-Video, our latest in #GenerativeAI research! With just a few words, this state-of-the-art AI system generates high-quality videos from text prompts.
— Meta AI (@MetaAI) September 29, 2022
Have an idea you want to see? Reply w/ your prompt using #MetaAI and we’ll share more results. pic.twitter.com/q8zjiwLBjb
As you may probably guess by the name, this is a new system that allows users to turn text into short-form video clips of the best quality as Zuckerberg bragged about in his blog on Thursday.
And the functioning is very similar to how Scene works, mixing different language processing and networks to use non-visual prompts to form pictures. In the end, it’s just like putting content into a new and unique format.
Meta says the idea behind the technology is very clear-cut and simple. It’s all about learning what the world appears like and how its best envisioned through the pairing of text with images. You also learn more about how it moves via unsupervised videos. This news came from the company’s team of researchers who published a paper on the findings of the new project.
When used correctly, the amount of time needed to have your video models trained is dramatically reduced. And it also gets rid of the need for paired text with video data sources too. All in all, you’ve got the diversity and depictions mastered brilliantly that can be found in the best generation models of today.
Just like similar projects coming forward from Meta’s AI team of researchers, this one is also going to be open-sourced. The company adds that it’s taking careful steps to build such innovative systems and share the results of the research with the world to harness feedback. They also vowed to better finetune and transform the approach to raging technology.
Just like any other AI release that comes with time, this opportunity for it to be misused can arise. And for this particular launch, it might not be a huge one. Hence, to rid any chances of malicious shenanigans, the team has eliminated any form of toxic phrasing and training dataset that comes along with it.
Read next: Meta Will Make History As It Plans Hiring Freeze, Reduced Headcount, And Team Reorganization For Facebook
