Downloads · 30 days
0
elismasilva/LTX-2.3-LoRa-Camera-Controls
LTX-2.3-LoRa-Camera-Controls is a text-to-video model from elismasilva. Use it when you need video from a text prompt. It is set up for ltx. The card lists the license as other.
A camera-control LoRA for the LTXV 2.3 base model. It adds support for camera movement phrases in the prompt, making it possible to guide video generations with tilt, pan, and zoom motion.
Downloads · 30 days
0
Access
Public
Updated Jul 9, 2026
Repo size
337 MB
Likes
3
Public
Click a slice to open those files.
.safetensors337 MB · 100%
From the Hugging Face model README
A camera-control LoRA for the LTXV 2.3 base model. It adds support for camera movement phrases in the prompt, making it possible to guide video generations with tilt, pan, and zoom motion.
Original reference: Camera Controls [LTX-2.3] on Civitai.
LTX2.3_CameraControls.safetensorsUse one of these phrases directly in the prompt to activate camera control:
the camera tilts up
the camera tilts down
the camera pans to the left
the camera pans to the right
the camera zooms in on <object>
the camera zooms out
For zoom-in prompts, replace <object> with the object or subject that should become the focus of the shot, for example:
the camera zooms in on the woman's face
Describe the scene and camera movement explicitly. For more consistent results, include:
Example:
A cinematic shot of a red sports car parked on a rainy city street, neon signs reflecting on the wet asphalt, the camera pans to the right revealing pedestrians and storefront lights.
Second example:
Prompt
A woman smiles at the camera and in an enthusiastic tone with a nordic accent says "L T X can now easily give camera directives. Watch now as the camera pans to the left" Once she's done speaking, the camera pans to the left continuously as she smiles and points that direction. As she speaks, her breath is visible on the cold air and the wind is blowing heavily causing flurries of snow to blow by.
Negative prompt
pc game, console game, video game, cartoon, childish, ugly
According to the Civitai model page, this LoRA was fine-tuned using synthetic clips. Version 1.0 used 188 clips after low-quality sets were removed, with the dataset created in 2026. The image data was processed through standard resizing, cropping, normalization, and labeling steps.
The creator states that the model is intended for non-commercial, experimental, and educational use. Users are responsible for complying with applicable copyright law, intellectual property law, and all other applicable laws.
See the terms published on the original model page: