LibTV Seedance 2.0 Live-action mode is now supported!
When Seedance 2.0 first came out, AI remix videos went viral on all major platforms—one-on-one on the basketball court with James, co-starring in a movie with Stephen Chow… It gave everyone their first taste of the terrifying power of AI video generation. But precisely because the results were so realistic, to avoid potential risks, within just a few days, Seedance 2.0 urgently restricted live-action material.
And this time on LibTV platform, not only has Seedance 2.0's live-action mode officially and compliantly returned, but generated videos also require no queueing at all。
We can have AI create a highly realistic selfie vlog anytime, anywhere.
Deliver a stand-up comedy routine with full emotion.
We can even directly generate coherent TV drama segments without actors.
Today, let's experience together how to use LibTV to create live-action videos convincing enough to pass as real!
01. LibTV Seedance 2.0 Live-Action Mode
We create a new project in LibTV.
Official Website：https://www.liblib.tv
To create live-action videos in LibTV, we need to use material content that has passed compliance verification, which specifically falls into two types:
AI Characters
We can directly upload AI-generated character images in LibTV, or directly use AI character images generated on the platform.
Right-click the image and select Seedance 2.0 Compliance Verification.
In just a few seconds, LibTV will complete the verification. When the small icon shown in the image below appears above the image, it means verification was successful, and the character can be used for Seedance 2.0 video generation.
Live-Action Portraits
If we want to use real-life human likenesses (such as ourselves or a hired model), the process is equally convenient.
We click on the character library in the video node.
Select live-action asset entry.
The system will generate an exclusive QR code, which requires the person to be generated to scan with their own phone to complete facial recognition detection and authorization confirmation.
After successful verification, we can add live-action images in this exclusive folder. After uploading images, the system will automatically review them. Note that the photos need to match the person who passed the liveness verification.
02. Hands-On Test
With compliant characters, we can call on these characters anytime to create live-action videos. I tried making several different types of cases, and the results were all very good.
The tiredness in the character's eyes is simply vivid and lifelike; tears fall on cue, and the facial muscle movements are also very natural—almost impossible to tell it's AI-generated.
The only minor flaw is the strands of hair stuck to the character's face; if they could also flutter in the wind, it would be even more realistic.
We input a complex dance instruction to test the model's physical performance.
I'm very satisfied with the generated result. Not only are the movements natural, but it also automatically generated very dynamic background music, and the character's dancing rhythm perfectly matches the musical beats.
I also casually made an electronic violin performance:
Seedance 2.0's synchronized audio-visual generation is truly so expressive—15 seconds is simply not enough!
The character's body movements and expressions are very realistic, completely free of the uncanny valley effect of early AI videos.
Now let's look at stand-up comedy:
The lip movements perfectly match the lines. The most impressive thing is that the character has a sense of ease; their gaze isn't emptily staring at the camera, but rather feels like they're interacting with the audience below the stage—very vivid.
Now let's look at a TV drama segment:
The character consistency is excellent. The two AI actors have very distinct personality traits—one is charming and lively, the other gentle and steady. It really has a TV drama quality to it.
03. Video production cost
The model performance is really impressive. Now let's talk about the production cost that everyone is more concerned about:
Currently, the most cost-effective option on the LibTV platform is the annual Creator membership, which is 39% off and also includes 180 free 15s video generations.
After calculating, the Seedance 2.0 VIP full-power version costs as little as 0.35 yuan/second, which is not only nearly half the price of the official website, but also completely queue-free.
If you are a professional-level user or team with long-term production needs and high demand, the membership supermarket can be used to top up Seedance 2.0 credits, with a limited-time buy-one-get-one-free offer.
04. Some sharing
After experiencing LibTV's live-action mode, my biggest feeling is: AI video finally has a live-action feel.
In the past, AI videos, no matter how high the image quality was, still had a strong AI flavor. The characters only had a good-looking face, and their movements and emotional expressions easily broke immersion.
This time, LibTV's live-action mode gives AI characters stable, controllable, and realistic expression. Many clips look even better than live-action quality. The problem of character consistency has also been solved, and complex movements are more natural.
AI live-action content is getting closer to our daily productivity. Content scenarios that highly depend on real people appearing on camera, such as advertising, e-commerce, short dramas, and personal IP, may soon be rewritten.
If we want to make live-action videos, we need to consider how to avoid the risk of portrait rights infringement and how to obtain compliant authorization. LibTV's insistence on launching compliance verification and QR code liveness authorization mechanisms is highly forward-looking. Only by truly solving the issues of authorization and security will creators dare to go all out to make money commercially.
Official website：https://www.liblib.tv
Original link:Seedance 2.0 lifts live-action restrictions! No queueing, as low as 0.35/second, come quickly!