MotionCLIP Text Encode

Documentation

Class name: MotionCLIPTextEncode
Category: MotionDiff
Output node: False

The MotionCLIPTextEncode node is designed to encode textual descriptions into a format suitable for motion generation tasks, leveraging a CLIP model and motion data to produce motion conditioning. This process facilitates the creation of motion sequences that are aligned with the given textual descriptions.

Input types

Required

md_clip
- The md_clip parameter represents a CLIP model tailored for motion data, crucial for interpreting and encoding the textual description in the context of motion.
- Comfy dtype: MD_CLIP
- Python dtype: MD_CLIP
motion_data
- The motion_data parameter contains the motion information necessary for the encoding process, providing the context in which the text is to be interpreted.
- Comfy dtype: MOTION_DATA
- Python dtype: MOTION_DATA
text
- The text parameter is the textual description to be encoded, serving as the input for generating motion-aligned sequences.
- Comfy dtype: STRING
- Python dtype: str

Output types

md_conditioning
- Comfy dtype: MD_CONDITIONING
- The output is motion conditioning data, which is used to guide the generation of motion sequences that correspond to the input text.
- Python dtype: MD_CONDITIONING

Usage tips

Infra type: GPU
Common nodes: unknown

Source code

class MotionCLIPTextEncode:
    @classmethod
    def INPUT_TYPES(s):
        return {
            "required": {
                "md_clip": ("MD_CLIP", ),
                "motion_data": ("MOTION_DATA", ),
                "text": ("STRING", {"default": "a person performs a cartwheel" ,"multiline": True})
            },
        }

    RETURN_TYPES = ("MD_CONDITIONING",)
    CATEGORY = "MotionDiff"
    FUNCTION = "encode_text"

    def encode_text(self, md_clip, motion_data, text):
        return (md_clip(text, motion_data), )

2.0 KiB Raw Permalink Blame History