nodes/chflame163-ComfyUI_LayerStyle_Advance
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Ultra nodes support vitmatte-base-composition-1k

Browse Source
This commit is contained in:
chflame163
2026-01-30 14:31:25 +08:00
parent c31de7030c
commit f16192bfa7
13 changed files with 47 additions and 30 deletions
Download Patch File Download Diff File Expand all files Collapse all files
py/ben_ultra.py
+2 -2
View File
@@ -991,7 +991,7 @@ class LS_BenUltra:
@classmethod
def INPUT_TYPES(cls):
method_list = ['VITMatte', 'VITMatte(local)', 'PyMatting', 'GuidedFilter', ]
method_list = ['VITMatte', 'VITMatte(local)', 'vitmatte-base-composition-1k', 'PyMatting', 'GuidedFilter', ]
return {
"required": {
"ben_model": ("BEN_MODEL",),
@@ -1044,7 +1044,7 @@ class LS_BenUltra:
_mask = tensor2pil(mask_edge_detail(i, _mask, detail_range // 8 + 1, black_point, white_point))
else:
_trimap = generate_VITMatte_trimap(_mask, detail_erode, detail_dilate)
_mask = generate_VITMatte(orig_image, _trimap, local_files_only=local_files_only, device=device, max_megapixels=max_megapixels)
_mask = generate_VITMatte(orig_image, _trimap, local_files_only=local_files_only, device=device, max_megapixels=max_megapixels, method=detail_method)
_mask = tensor2pil(histogram_remap(pil2tensor(_mask), black_point, white_point))
else:
_mask = tensor2pil(_mask)
py/birefnet_ultra.py
+2 -2
View File
@@ -9,7 +9,7 @@ class BiRefNetUltra:
@classmethod
def INPUT_TYPES(cls):
method_list = ['VITMatte', 'VITMatte(local)', 'PyMatting', 'GuidedFilter', ]
method_list = ['VITMatte', 'VITMatte(local)', 'vitmatte-base-composition-1k', 'PyMatting', 'GuidedFilter', ]
device_list = ['cuda','cpu']
return {
"required": {
@@ -62,7 +62,7 @@ class BiRefNetUltra:
_mask = tensor2pil(mask_edge_detail(i, _mask, detail_range // 8 + 1, black_point, white_point))
else:
_trimap = generate_VITMatte_trimap(_mask, detail_erode, detail_dilate)
_mask = generate_VITMatte(orig_image, _trimap, local_files_only=local_files_only, device=device, max_megapixels=max_megapixels)
_mask = generate_VITMatte(orig_image, _trimap, local_files_only=local_files_only, device=device, max_megapixels=max_megapixels, method=detail_method)
_mask = tensor2pil(histogram_remap(pil2tensor(_mask), black_point, white_point))
else:
_mask = tensor2pil(_mask)
py/birefnet_ultra_v2.py
+2 -2
View File
@@ -112,7 +112,7 @@ class LS_BiRefNetUltraV2:
@classmethod
def INPUT_TYPES(cls):
method_list = ['VITMatte', 'VITMatte(local)', 'PyMatting', 'GuidedFilter', ]
method_list = ['VITMatte', 'VITMatte(local)', 'vitmatte-base-composition-1k', 'PyMatting', 'GuidedFilter', ]
device_list = ['cuda', 'cpu']
return {
"required": {
@@ -187,7 +187,7 @@ class LS_BiRefNetUltraV2:
_mask = tensor2pil(mask_edge_detail(i, _mask, detail_range // 8 + 1, black_point, white_point))
else:
_trimap = generate_VITMatte_trimap(_mask, detail_erode, detail_dilate)
_mask = generate_VITMatte(orig_image, _trimap, local_files_only=local_files_only, device=device, max_megapixels=max_megapixels)
_mask = generate_VITMatte(orig_image, _trimap, local_files_only=local_files_only, device=device, max_megapixels=max_megapixels, method=detail_method)
_mask = tensor2pil(histogram_remap(pil2tensor(_mask), black_point, white_point))
else:
_mask = tensor2pil(_mask)
py/evf_sam_ultra.py
+2 -2
View File
@@ -20,7 +20,7 @@ class EVF_SAM_Ultra:
model_list = ["evf-sam2", "evf-sam"]
precision_list = ["fp16", "bf16", "fp32"]
load_in_bit_list = ["full", "8", "4"]
method_list = ['VITMatte', 'VITMatte(local)', 'PyMatting', 'GuidedFilter', ]
method_list = ['VITMatte', 'VITMatte(local)', 'vitmatte-base-composition-1k', 'PyMatting', 'GuidedFilter', ]
device_list = ['cuda', 'cpu']
return {"required":
{
@@ -98,7 +98,7 @@ class EVF_SAM_Ultra:
else:
_trimap = generate_VITMatte_trimap(_mask, detail_erode, detail_dilate)
_mask = generate_VITMatte(orig_image, _trimap, local_files_only=local_files_only, device=device,
max_megapixels=max_megapixels)
max_megapixels=max_megapixels, method=detail_method)
_mask = tensor2pil(histogram_remap(pil2tensor(_mask), black_point, white_point))
else:
_mask = mask2image(_mask)
py/florence2_ultra.py
+2 -2
View File
@@ -457,7 +457,7 @@ class Florence2Ultra:
"referring expression segmentation",
"open vocabulary detection",
]
method_list = ['VITMatte', 'VITMatte(local)', 'PyMatting', 'GuidedFilter', ]
method_list = ['VITMatte', 'VITMatte(local)', 'vitmatte-base-composition-1k', 'PyMatting', 'GuidedFilter', ]
device_list = ['cuda','cpu']
return {
"required": {
@@ -522,7 +522,7 @@ class Florence2Ultra:
_mask = tensor2pil(mask_edge_detail(i, _mask, detail_range // 8 + 1, black_point, white_point))
else:
_trimap = generate_VITMatte_trimap(_mask, detail_erode, detail_dilate)
_mask = generate_VITMatte(img, _trimap, local_files_only=local_files_only, device=device, max_megapixels=max_megapixels)
_mask = generate_VITMatte(img, _trimap, local_files_only=local_files_only, device=device, max_megapixels=max_megapixels, method=detail_method)
_mask = tensor2pil(histogram_remap(pil2tensor(_mask), black_point, white_point))
else:
_mask = tensor2pil(_mask)
py/human_parts_ultra.py
+2 -2
View File
@@ -34,7 +34,7 @@ class LS_HumanPartsUltra:
@classmethod
def INPUT_TYPES(cls):
method_list = ['VITMatte', 'VITMatte(local)', 'PyMatting', 'GuidedFilter', ]
method_list = ['VITMatte', 'VITMatte(local)', 'vitmatte-base-composition-1k', 'PyMatting', 'GuidedFilter', ]
device_list = ['cuda', 'cpu']
return {
"required": {
@@ -104,7 +104,7 @@ class LS_HumanPartsUltra:
else:
_trimap = generate_VITMatte_trimap(_mask, detail_erode, detail_dilate)
_mask = generate_VITMatte(orig_image, _trimap, local_files_only=local_files_only, device=device,
max_megapixels=max_megapixels)
max_megapixels=max_megapixels, method=detail_method)
_mask = tensor2pil(histogram_remap(pil2tensor(_mask), black_point, white_point))
else:
_mask = mask2image(_mask)
py/imagefunc.py
+20 -3
View File
@@ -1569,7 +1569,19 @@ def load_VITMatte_model(model_name:str, local_files_only:bool=False) -> object:
vitmatte = VITMatteModel(model, processor)
return vitmatte
def generate_VITMatte(image:Image, trimap:Image, local_files_only:bool=False, device:str="cpu", max_megapixels:float=2.0) -> Image:
def load_VITMatte_base_model(model_name:str, local_files_only:bool=False) -> object:
model_name = "vitmatte-base-composition-1k"
model_repo = "hustvl/vitmatte-base-composition-1k"
model_path = check_and_download_model(model_name, model_repo)
from transformers import VitMatteImageProcessor, VitMatteForImageMatting
model = VitMatteForImageMatting.from_pretrained(model_path, local_files_only=local_files_only)
processor = VitMatteImageProcessor.from_pretrained(model_path, local_files_only=local_files_only)
vitmatte = VITMatteModel(model, processor)
return vitmatte
def generate_VITMatte(image:Image, trimap:Image, local_files_only:bool=False, device:str="cpu",
max_megapixels:float=2.0, method:str="VITMatte") -> Image:
if image.mode != 'RGB':
image = image.convert('RGB')
if trimap.mode != 'L':
@@ -1585,7 +1597,7 @@ def generate_VITMatte(image:Image, trimap:Image, local_files_only:bool=False, de
image = image.resize((target_width, target_height), Image.BILINEAR)
trimap = trimap.resize((target_width, target_height), Image.BILINEAR)
# log(f"vitmatte image size {width}x{height} too large, resize to {target_width}x{target_height} for processing.")
model_name = "hustvl/vitmatte-small-composition-1k"
if device=="cpu":
device = torch.device('cpu')
else:
@@ -1594,7 +1606,12 @@ def generate_VITMatte(image:Image, trimap:Image, local_files_only:bool=False, de
else:
log("vitmatte device is set to cuda, but not available, using cpu instead.")
device = torch.device('cpu')
vit_matte_model = load_VITMatte_model(model_name=model_name, local_files_only=local_files_only)
if method == "vitmatte-base-composition-1k":
model_name = "hustvl/vitmatte-base-composition-1k"
vit_matte_model = load_VITMatte_base_model(model_name=model_name, local_files_only=local_files_only)
else:
model_name = "hustvl/vitmatte-small-composition-1k"
vit_matte_model = load_VITMatte_model(model_name=model_name, local_files_only=local_files_only)
vit_matte_model.model.to(device)
# log(f"vitmatte processing, image size = {image.width}x{image.height}, device = {device}.")
inputs = vit_matte_model.processor(images=image, trimaps=trimap, return_tensors="pt")
py/person_mask_ultra_v2.py
+2 -2
View File
@@ -19,7 +19,7 @@ class PersonMaskUltraV2:
@classmethod
def INPUT_TYPES(self):
method_list = ['VITMatte', 'VITMatte(local)', 'PyMatting', 'GuidedFilter', ]
method_list = ['VITMatte', 'VITMatte(local)', 'vitmatte-base-composition-1k', 'PyMatting', 'GuidedFilter', ]
device_list = ['cuda','cpu']
return {
"required":
@@ -162,7 +162,7 @@ class PersonMaskUltraV2:
detail_range // 8 + 1, black_point, white_point))
else:
_trimap = generate_VITMatte_trimap(_mask, detail_erode, detail_dilate)
_mask = generate_VITMatte(orig_image, _trimap, local_files_only=local_files_only, device=device, max_megapixels=max_megapixels)
_mask = generate_VITMatte(orig_image, _trimap, local_files_only=local_files_only, device=device, max_megapixels=max_megapixels, method=detail_method)
_mask = tensor2pil(histogram_remap(pil2tensor(_mask), black_point, white_point))
else:
_mask = mask2image(_mask)
py/sam_2_ultra.py
+6 -6
View File
@@ -230,7 +230,7 @@ class LS_SAM2_ULTRA:
]
model_precision_list = [ 'fp16','bf16','fp32']
select_list = ["all", "first", "by_index"]
method_list = ['VITMatte', 'VITMatte(local)', 'PyMatting', 'GuidedFilter', ]
method_list = ['VITMatte', 'VITMatte(local)', 'vitmatte-base-composition-1k', 'PyMatting', 'GuidedFilter', ]
device_list = ['cuda','cpu']
return {
"required": {
@@ -419,7 +419,7 @@ class LS_SAM2_ULTRA:
else:
_trimap = generate_VITMatte_trimap(_mask, detail_erode, detail_dilate)
_mask = generate_VITMatte(orig_image, _trimap, local_files_only=local_files_only, device=device,
max_megapixels=max_megapixels)
max_megapixels=max_megapixels, method=detail_method)
_mask = tensor2pil(histogram_remap(pil2tensor(_mask), black_point, white_point))
else:
_mask = tensor2pil(_mask)
@@ -552,7 +552,7 @@ class LS_SAM2_ULTRA_V2:
model_precision_list = [ 'fp16','bf16','fp32']
select_list = ["all", "first", "by_index"]
method_list = ['VITMatte', 'VITMatte(local)', 'PyMatting', 'GuidedFilter', ]
method_list = ['VITMatte', 'VITMatte(local)', 'vitmatte-base-composition-1k', 'PyMatting', 'GuidedFilter', ]
device_list = ['cuda','cpu']
return {
"required": {
@@ -689,7 +689,7 @@ class LS_SAM2_ULTRA_V2:
else:
_trimap = generate_VITMatte_trimap(_mask, detail_erode, detail_dilate)
_mask = generate_VITMatte(orig_image, _trimap, local_files_only=local_files_only, device=device,
max_megapixels=max_megapixels)
max_megapixels=max_megapixels, method=detail_method)
_mask = tensor2pil(histogram_remap(pil2tensor(_mask), black_point, white_point))
else:
_mask = tensor2pil(_mask)
@@ -751,7 +751,7 @@ class LS_SAM2_VIDEO_ULTRA:
'sam2.1_hiera_tiny.safetensors',
]
model_precision_list = ['fp16','bf16']
method_list = ['VITMatte']
method_list = ['VITMatte', 'vitmatte-base-composition-1k', ]
device_list = ['cuda']
return {
"required": {
@@ -1029,7 +1029,7 @@ class LS_SAM2_VIDEO_ULTRA:
if process_detail:
_trimap = generate_VITMatte_trimap(pil2tensor(_mask), detail_erode, detail_dilate)
_mask = generate_VITMatte(orig_image, _trimap, local_files_only=local_files_only, device=device,
max_megapixels=max_megapixels)
max_megapixels=max_megapixels, method=detail_method)
_mask = tensor2pil(histogram_remap(pil2tensor(_mask), black_point, white_point))
color_image = Image.new("RGB", orig_image.size,color=mask_preview_color)
py/segment_anything_ultra_v2.py
+2 -2
View File
@@ -19,7 +19,7 @@ class SegmentAnythingUltraV2:
@classmethod
def INPUT_TYPES(cls):
method_list = ['VITMatte', 'VITMatte(local)', 'PyMatting', 'GuidedFilter', ]
method_list = ['VITMatte', 'VITMatte(local)', 'vitmatte-base-composition-1k', 'PyMatting', 'GuidedFilter', ]
device_list = ['cuda','cpu']
return {
"required": {
@@ -88,7 +88,7 @@ class SegmentAnythingUltraV2:
_mask = tensor2pil(mask_edge_detail(i, _mask, detail_range // 8 + 1, black_point, white_point))
else:
_trimap = generate_VITMatte_trimap(_mask, detail_erode, detail_dilate)
_mask = generate_VITMatte(_image, _trimap, local_files_only=local_files_only, device=device, max_megapixels=max_megapixels)
_mask = generate_VITMatte(_image, _trimap, local_files_only=local_files_only, device=device, max_megapixels=max_megapixels, method=detail_method)
_mask = tensor2pil(histogram_remap(pil2tensor(_mask), black_point, white_point))
else:
_mask = mask2image(_mask)
py/segment_anything_ultra_v3.py
+2 -2
View File
@@ -39,7 +39,7 @@ class LS_SegmentAnythingUltraV3:
@classmethod
def INPUT_TYPES(cls):
method_list = ['VITMatte', 'VITMatte(local)', 'PyMatting', 'GuidedFilter', ]
method_list = ['VITMatte', 'VITMatte(local)', 'vitmatte-base-composition-1k', 'PyMatting', 'GuidedFilter', ]
device_list = ['cuda','cpu']
return {
"required": {
@@ -102,7 +102,7 @@ class LS_SegmentAnythingUltraV3:
_mask = tensor2pil(mask_edge_detail(i, _mask, detail_range // 8 + 1, black_point, white_point))
else:
_trimap = generate_VITMatte_trimap(_mask, detail_erode, detail_dilate)
_mask = generate_VITMatte(_image, _trimap, local_files_only=local_files_only, device=device, max_megapixels=max_megapixels)
_mask = generate_VITMatte(_image, _trimap, local_files_only=local_files_only, device=device, max_megapixels=max_megapixels, method=detail_method)
_mask = tensor2pil(histogram_remap(pil2tensor(_mask), black_point, white_point))
else:
_mask = mask2image(_mask)
py/transparent_background_ultra.py
+2 -2
View File
@@ -19,7 +19,7 @@ class TransparentBackgroundUltra:
@classmethod
def INPUT_TYPES(cls):
method_list = ['VITMatte', 'VITMatte(local)', 'PyMatting', 'GuidedFilter', ]
method_list = ['VITMatte', 'VITMatte(local)', 'vitmatte-base-composition-1k', 'PyMatting', 'GuidedFilter', ]
device_list = ['cuda','cpu']
return {
@@ -78,7 +78,7 @@ class TransparentBackgroundUltra:
_mask = tensor2pil(mask_edge_detail(i, _mask, detail_range // 8 + 1, black_point, white_point))
else:
_trimap = generate_VITMatte_trimap(_mask, detail_erode, detail_dilate)
_mask = generate_VITMatte(orig_image, _trimap, local_files_only=local_files_only, device=device, max_megapixels=max_megapixels)
_mask = generate_VITMatte(orig_image, _trimap, local_files_only=local_files_only, device=device, max_megapixels=max_megapixels, method=detail_method)
_mask = tensor2pil(histogram_remap(pil2tensor(_mask), black_point, white_point))
ret_image = RGB2RGBA(orig_image, _mask.convert('L'))
pyproject.toml
+1 -1
View File
@@ -1,7 +1,7 @@
[project]
name = "ComfyUI_LayerStyle_Advance"
description = "The nodes detached from ComfyUI Layer Style are mainly those with complex requirements for dependency packages."
version = "2.0.35"
version = "2.0.36"
license = { text = "MIT License" }
dependencies = ["numpy", "matplotlib", "scikit_image", "scikit_learn", "opencv-contrib-python", "pymatting", "timm", "blend_modes", "transformers", "diffusers", "loguru", "colour-science", "huggingface_hub", "segment_anything", "addict", "omegaconf", "yapf", "wget", "iopath", "mediapipe", "typer_config", "fastapi", "rich", "google-generativeai", "ultralytics", "transparent-background", "accelerate", "onnxruntime", "bitsandbytes", "peft", "protobuf", "hydra-core", "blind-watermark", "qrcode", "pyzbar", "psd-tools", "wandb", "zhipuai", "openai","google-genai", "fastapi","typer-config"]