152 lines
7.0 KiB
Python
152 lines
7.0 KiB
Python
from .utils import runwareUtils as rwUtils
|
|
from typing import List
|
|
|
|
|
|
class videoInferenceInputs:
|
|
"""Video Inference Inputs node for configuring video generation inputs"""
|
|
|
|
@classmethod
|
|
def INPUT_TYPES(cls):
|
|
return {
|
|
"required": {},
|
|
"optional": {
|
|
"Image": ("IMAGE", {
|
|
"tooltip": "Portrait image for video generation (main subject that will speak and move). Mutually exclusive with Avatar: when Avatar is set, Image is ignored."
|
|
}),
|
|
"Avatar": ("STRING", {
|
|
"default": "",
|
|
"tooltip": "HeyGen avatar ID for a photo or video avatar. Mutually exclusive with Image: when set, Image is ignored; use one or the other.",
|
|
}),
|
|
"Frame Images": ("RUNWAREVIDEOINPUTSFRAMEIMAGES", {
|
|
"tooltip": "Connect the Runware Video Inputs Frame node to provide timeline-constrained frame images."
|
|
}),
|
|
"Audio": ("STRING", {
|
|
"tooltip": "Connect the mediaUUID output from Runware Media Upload node with audio file. This audio will be synchronized with the generated video. (Legacy: single audio string)"
|
|
}),
|
|
"Audio Inputs": ("RUNWAREAUDIOINPUT", {
|
|
"tooltip": "Connect Runware Video Audio Input node(s) to provide audio inputs with reference IDs."
|
|
}),
|
|
"Speech Inputs": ("RUNWARESPEECHINPUT", {
|
|
"tooltip": "Connect Runware Video Speech Input node or Runware Speech Inputs Combine node to provide TTS speech inputs with reference IDs."
|
|
}),
|
|
"Video": ("STRING", {
|
|
"tooltip": "Connect the mediaUUID output from Runware Media Upload node with a reference/input video."
|
|
}),
|
|
"Mask": ("IMAGE", {
|
|
"tooltip": "Mask image to specify a specific subject in the image to speak. Use white/black mask format."
|
|
}),
|
|
"Frame": ("IMAGE", {
|
|
"tooltip": "Frame image for video generation. Connect a Load Image node to provide the frame image."
|
|
}),
|
|
"Background": ("IMAGE", {
|
|
"tooltip": "Background image for video generation. Connect a Load Image node to provide the background.",
|
|
}),
|
|
"References": ("RUNWAREVIDEOINPUTSREFERENCEIMAGES", {
|
|
"tooltip": "Connect the Video Inputs References node to provide reference images."
|
|
}),
|
|
"Reference Videos": ("RUNWAREREFERENCEVIDEOS", {
|
|
"tooltip": "Connect the Reference Videos node to provide reference video mediaUUIDs."
|
|
}),
|
|
"Reference Voices": ("RUNWAREREFERENCEVOICES", {
|
|
"tooltip": "Connect the Reference Voices node to provide reference voice URLs or mediaUUIDs."
|
|
}),
|
|
"draftId": ("STRING", {
|
|
"default": "",
|
|
"multiline": False,
|
|
"tooltip": "Draft ID from a previous Bytedance video generation. Used when continuing or extending a draft output",
|
|
}),
|
|
"videoId": ("STRING", {
|
|
"default": "",
|
|
"multiline": False,
|
|
"tooltip": "Video ID from a previous generation. Used when extending or editing an existing OpenAI Sora video",
|
|
}),
|
|
}
|
|
}
|
|
|
|
DESCRIPTION = (
|
|
"Configure custom inputs for Runware Video Inference (image, audio, video, mask, avatar, etc.). "
|
|
"Avatar and Image are mutually exclusive: if Avatar is set, Image is ignored; provide only one."
|
|
)
|
|
FUNCTION = "createInputs"
|
|
RETURN_TYPES = ("RUNWAREVIDEOINFERENCEINPUTS",)
|
|
RETURN_NAMES = ("Video Inference Inputs",)
|
|
CATEGORY = "Runware"
|
|
|
|
def createInputs(self, **kwargs):
|
|
"""Create video inference inputs from provided parameters"""
|
|
image = kwargs.get("Image", None)
|
|
audio = kwargs.get("Audio", None)
|
|
audioInputs = kwargs.get("Audio Inputs", None)
|
|
speechInputs = kwargs.get("Speech Inputs", None)
|
|
video = kwargs.get("Video", None)
|
|
mask = kwargs.get("Mask", None)
|
|
frame = kwargs.get("Frame", None)
|
|
background = kwargs.get("Background", None)
|
|
frameImages = kwargs.get("Frame Images", None)
|
|
references = kwargs.get("References", None)
|
|
referenceVideos = kwargs.get("Reference Videos", None)
|
|
referenceVoices = kwargs.get("Reference Voices", None)
|
|
draftId = kwargs.get("draftId", "")
|
|
videoId = kwargs.get("videoId", "")
|
|
avatar = kwargs.get("Avatar", "")
|
|
|
|
inputs = {}
|
|
|
|
has_avatar = avatar is not None and avatar.strip() != ""
|
|
if has_avatar:
|
|
inputs["avatar"] = avatar.strip()
|
|
if image is not None and not has_avatar:
|
|
inputs["image"] = rwUtils.convertTensor2IMG(image)
|
|
|
|
# Handle audio: can be either a string (legacy), a single dict, or a list of audio inputs
|
|
if audioInputs is not None:
|
|
if isinstance(audioInputs, list) and len(audioInputs) > 0:
|
|
# List from combine node
|
|
inputs["audio"] = audioInputs
|
|
elif isinstance(audioInputs, dict) and len(audioInputs) > 0:
|
|
# Single audio input dict
|
|
inputs["audio"] = [audioInputs]
|
|
elif audio is not None and audio.strip() != "":
|
|
inputs["audio"] = audio.strip()
|
|
|
|
# Handle speech inputs: can be a single dict or a list of speech inputs
|
|
if speechInputs is not None:
|
|
if isinstance(speechInputs, list) and len(speechInputs) > 0:
|
|
# List from combine node
|
|
inputs["speech"] = speechInputs
|
|
elif isinstance(speechInputs, dict) and len(speechInputs) > 0:
|
|
# Single speech input dict
|
|
inputs["speech"] = [speechInputs]
|
|
|
|
if video is not None and video.strip() != "":
|
|
inputs["video"] = video.strip()
|
|
|
|
if mask is not None:
|
|
inputs["mask"] = rwUtils.convertTensor2IMG(mask)
|
|
|
|
if frame is not None:
|
|
inputs["frame"] = rwUtils.convertTensor2IMG(frame)
|
|
|
|
if background is not None:
|
|
inputs["background"] = rwUtils.convertTensor2IMG(background)
|
|
|
|
if frameImages is not None and len(frameImages) > 0:
|
|
inputs["frameImages"] = frameImages
|
|
|
|
if references is not None and len(references) > 0:
|
|
inputs["referenceImages"] = references
|
|
|
|
if referenceVideos is not None and len(referenceVideos) > 0:
|
|
inputs["referenceVideos"] = referenceVideos
|
|
|
|
if referenceVoices is not None and len(referenceVoices) > 0:
|
|
inputs["referenceVoices"] = referenceVoices
|
|
|
|
if draftId is not None and draftId.strip() != "":
|
|
inputs["draftId"] = draftId.strip()
|
|
|
|
if videoId is not None and videoId.strip() != "":
|
|
inputs["videoId"] = videoId.strip()
|
|
|
|
return (inputs,)
|