Superside ComfyUI NodesandClaude Sonnet 5
15c80a8335
New SupersideImageRetouchNode: one-click retouch/clean-up of an image (skin, blemishes, imperfections) via fal's image-editing retouch model. No prompt needed - takes an image plus optional guidance_scale, num_inference_steps, lora_scale (retouch strength), seed, safety checker, and sync_mode; returns the retouched image. Verified end-to-end against the live fal API. Registered and documented in the README. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
124 lines
4.6 KiB
Python
124 lines
4.6 KiB
Python
import logging
|
|
|
|
from .base_node import (
|
|
SupersideFalNode,
|
|
ImageProcessingMixin,
|
|
APIClientMixin,
|
|
API_KEY_INPUT_SPEC,
|
|
)
|
|
|
|
logger = logging.getLogger(__name__)
|
|
|
|
|
|
class SupersideImageRetouchNode(
|
|
SupersideFalNode, ImageProcessingMixin, APIClientMixin
|
|
):
|
|
"""
|
|
Image Retouch Node: retouch/clean up an image (e.g. skin, blemishes,
|
|
imperfections) using fal.ai's image-editing retouch model
|
|
(endpoint "fal-ai/image-editing/retouch").
|
|
|
|
Takes a single image and returns a retouched version. No prompt needed -
|
|
the model applies its retouching pass automatically.
|
|
"""
|
|
|
|
@classmethod
|
|
def INPUT_TYPES(cls):
|
|
return {
|
|
"required": {
|
|
"image": ("IMAGE",),
|
|
"api_key": API_KEY_INPUT_SPEC,
|
|
},
|
|
"optional": {
|
|
"guidance_scale": (
|
|
"FLOAT",
|
|
{"default": 3.5, "min": 0.0, "max": 20.0, "step": 0.1,
|
|
"tooltip": "CFG scale - how strongly the model follows its retouch objective. Higher = stronger effect."},
|
|
),
|
|
"num_inference_steps": (
|
|
"INT",
|
|
{"default": 30, "min": 1, "max": 100,
|
|
"tooltip": "Number of sampling steps. More steps = potentially cleaner result, slower."},
|
|
),
|
|
"lora_scale": (
|
|
"FLOAT",
|
|
{"default": 1.0, "min": 0.0, "max": 2.0, "step": 0.05,
|
|
"tooltip": "Strength of the retouch LoRA. Lower = subtler retouch, higher = stronger."},
|
|
),
|
|
"seed": ("INT", {"default": -1, "min": -1, "max": 2147483647, "tooltip": "-1 = random"}),
|
|
"enable_safety_checker": ("BOOLEAN", {"default": True}),
|
|
"sync_mode": ("BOOLEAN", {"default": False}),
|
|
},
|
|
"hidden": {
|
|
"unique_id": "UNIQUE_ID",
|
|
"extra_pnginfo": "EXTRA_PNGINFO",
|
|
},
|
|
}
|
|
|
|
RETURN_TYPES = ("IMAGE", "STRING")
|
|
RETURN_NAMES = ("image", "info")
|
|
FUNCTION = "generate"
|
|
OUTPUT_NODE = True
|
|
DESCRIPTION = (
|
|
"Retouch/clean up an image (skin, blemishes, imperfections) using "
|
|
"fal.ai's image-editing retouch model (fal-ai/image-editing/retouch). "
|
|
"No prompt needed - just connect an image."
|
|
)
|
|
|
|
def prepare_arguments(self, client, image, **kwargs):
|
|
image_url = self.upload_image(client, image)
|
|
arguments = {"image_url": image_url}
|
|
|
|
if kwargs.get("guidance_scale") is not None:
|
|
arguments["guidance_scale"] = kwargs["guidance_scale"]
|
|
|
|
if kwargs.get("num_inference_steps") is not None:
|
|
arguments["num_inference_steps"] = kwargs["num_inference_steps"]
|
|
|
|
if kwargs.get("lora_scale") is not None:
|
|
arguments["lora_scale"] = kwargs["lora_scale"]
|
|
|
|
seed = kwargs.get("seed", -1)
|
|
if seed is not None and seed != -1:
|
|
arguments["seed"] = seed
|
|
|
|
if kwargs.get("enable_safety_checker") is not None:
|
|
arguments["enable_safety_checker"] = kwargs["enable_safety_checker"]
|
|
|
|
if kwargs.get("sync_mode") is not None:
|
|
arguments["sync_mode"] = kwargs["sync_mode"]
|
|
|
|
return arguments
|
|
|
|
def generate(self, image, api_key, unique_id=None, extra_pnginfo=None, **kwargs):
|
|
try:
|
|
client = self.get_client(api_key)
|
|
arguments = self.prepare_arguments(client, image, **kwargs)
|
|
result = self.call_api(client, "fal-ai/image-editing/retouch", arguments)
|
|
|
|
images = self.process_images(result)
|
|
|
|
first_url = ""
|
|
if isinstance(result.get("images"), list) and result["images"]:
|
|
first_url = result["images"][0].get("url", "")
|
|
info = first_url
|
|
|
|
if unique_id is not None and extra_pnginfo is not None:
|
|
if (
|
|
isinstance(extra_pnginfo, list)
|
|
and isinstance(extra_pnginfo[0], dict)
|
|
and "workflow" in extra_pnginfo[0]
|
|
):
|
|
workflow = extra_pnginfo[0]["workflow"]
|
|
node = next(
|
|
(x for x in workflow["nodes"] if str(x["id"]) == str(unique_id)),
|
|
None,
|
|
)
|
|
if node:
|
|
node["widgets_values"] = [info]
|
|
|
|
return {"ui": {"text": [info]}, "result": (images[0], info)}
|
|
except Exception as e:
|
|
logger.error(f"Image retouch failed: {str(e)}")
|
|
raise RuntimeError(f"Image retouch failed: {str(e)}") from e
|