Creates the famous "Dolly Zoom" cinematic effect.
https://en.wikipedia.org/wiki/Dolly_zoom
I am going to revisit this model after I have collected more training data. You need to lower the strength (0.8) in order for it to play well with character models. I would prefer it not bias towards specific faces.
a dramatic dolly-zoom shot
in the foreground is a large lemur that is center frame, the lemur is holding a smaller lemur in her hands and is looking at the camera
the larger lemur wearing a sleek white minidress with a high neckline
in the background is a hallway that stretches into the distance
bright lighting
dolly-zoom shot where the background shiftsDescription
FAQ
Comments (7)
This is perfect. I really hope we see more of these types of loras for Hunyuan. cant wait to try it out today.
Finally something SFW! Please keep it up! Hope to see more from you, take some buzz!
Lemurs having a sudden realisation that they left the gas on at home is the best use of AI so far!
YES! thank you!
Great idea. Can you please do other camera movement loras? Like a simple regular zoom, and a simple regular dolly? Base Hunyuan isn't very good at following basic camera instructions like that.
And more advanced: Zoom or dolly to a specific object, rather than the center
I tested this and it work real well. We need more tooling like this. Once I finish my next video I will be doing some of these. great work! will add clips latter today.
Using this lora in framepack (studio) completely changes the experience. This lora can be used to change scenes and background more easy or create morph effects between timestamped prompts, and as a bonus it does dolly zoom on characters on the video
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.