- Luma Labs’ new modifying video tool for Dream Machine uses AI to modify all video sequences without recovery
- All characters or environments will not lose their original movement or performance
- All that, from subtle adjustments to the wardrobe, to the revision of the complete magic scene, is possible
Luma Labs is known to produce AI videos from zero, but society has a new feature for its dream machine which can completely transform real video sequences in a subtle or flagrant way, even if it is only an old home film.
The new Modify video feature makes something like the best Photoshop tools for images for videos. He can modify the parameter, the style of a scene, even the whole characters, all without restart, resuscitation or even stand up.
The company boasts that the video edition of the AI retains everything that matters to you from the original recording, such as the movement of the actors, the framing, the timing and other key details, while modifying everything you want.
The outfit you wear, what you have decided was not you, is suddenly a whole set of fully different clothes. This cover fort is now a ship sailing in a stormy sea, and your friend who rushes on the ground is actually an astronaut in space, all without using green screens or bay editing.
The combination of advanced movement luma and performance capture, IA style and what he calls structured presets makes it possible to offer the entire range of redesigned videos.
All you have to do is download a video up to 10 seconds to start. Then choose from the members, flexible or reinvented.
Adhering is the most subtle option; It focuses on minimum changes, such as adjustment of the clothes below or different textures on the furniture. Flex does it, but can also adjust the style of the video, lighting and other more obvious details. The reimagine, as its name suggests, can refuse everything on the video, take it to another world or redo people in animal animals or send someone standing on a flat board in a cyberpunk hoverboard breed.
Ai flexible video
It all depends not only on the prompts, but the reference images and the frame selections from your video if you choose. As a result, the process is much more user -friendly and flexible.
Although the AI video modification is hardly unique to Luma, the company claims that it surpasses the rivals like Runway and Pika because of its performance loyalty. Modified videos keep the body language of an actor, facial expressions and lip synchronization. The final results appear as an organic whole, not just sewn bits.
Of course, Modify video tools have limits. These are always capped at 10 seconds per clip for the moment, which maintains the manageable things in terms of waiting time. However, if you want a longer film, you need to plan and determine how to artistically incorporate different plans into a single film.
However, features such as the possibility of isolating the elements in a shot are a big problem. Sometimes you have a performance of which you are very satisfied, but it is supposed to be another type of character in a different setting. Well, you can keep the performance intact and exchange a garage against the sea and the legs of your actor for a fish tail.
Dreams to reality
It is really impressive to see how quickly and deep AI tools can rework a little sequences. These tools are not only a gadget; AI models are aware of performance and deadlines in a way that feels closer to man than anything I have seen. AI models do not really understand stimulation, continuity or structure, but they are very good at imitating these aspects.
Although the technical and ethical limitations will prevent Luma Labs from recreating all the cinema at this stage, these tools will be tempted for many producers of amateur or independent videos. And even if I do not see that it also becomes widely used as common photo filters, there are fun ideas in Luma demos that you may want to try.
