Skip to main content
This documentation was AI-generated. If you find any errors or have suggestions for improvement, please feel free to contribute! Edit on GitHub
The WanCameraEmbedding node generates camera trajectory embeddings using Plücker embeddings based on camera motion parameters. It creates a sequence of camera poses that simulate different camera movements and converts them into embedding tensors suitable for video generation pipelines.

Inputs

ParameterData TypeRequiredRangeDescription
camera_poseCOMBOYes”Static"
"Pan Up"
"Pan Down"
"Pan Left"
"Pan Right"
"Zoom In"
"Zoom Out"
"Anti Clockwise (ACW)"
"ClockWise (CW)“
The type of camera movement to simulate (default: “Static”)
widthINTYes16 to MAX_RESOLUTIONThe width of the output in pixels (default: 832, step: 16)
heightINTYes16 to MAX_RESOLUTIONThe height of the output in pixels (default: 480, step: 16)
lengthINTYes1 to MAX_RESOLUTIONThe length of the camera trajectory sequence (default: 81, step: 4)
speedFLOATNo0.0 to 10.0The speed of the camera movement (default: 1.0, step: 0.1)
fxFLOATNo0.0 to 1.0The focal length x parameter (default: 0.5, step: 0.000000001)
fyFLOATNo0.0 to 1.0The focal length y parameter (default: 0.5, step: 0.000000001)
cxFLOATNo0.0 to 1.0The principal point x coordinate (default: 0.5, step: 0.01)
cyFLOATNo0.0 to 1.0The principal point y coordinate (default: 0.5, step: 0.01)

Outputs

Output NameData TypeDescription
camera_embeddingTENSORThe generated camera embedding tensor containing the trajectory sequence
widthINTThe width value that was used for processing
heightINTThe height value that was used for processing
lengthINTThe length value that was used for processing