app.py

# app.py

# ====== PATCH GRADIO_CLIENT UTILS TO HANDLE BOOLEAN SCHEMAS ======
import gradio_client.utils as gc_utils

def patched_get_type(schema):
    if not isinstance(schema, dict):
        return "bool" if isinstance(schema, bool) else "unknown"
    if "const" in schema:
        return "const"
    return schema.get("type", "object")
gc_utils.get_type = patched_get_type

_original_json_schema_to_python_type = gc_utils._json_schema_to_python_type
def patched_json_schema_to_python_type(schema, defs=None):
    if isinstance(schema, bool):
        return "bool"
    if not isinstance(schema, dict):
        return "unknown"
    try:
        return _original_json_schema_to_python_type(schema, defs)
    except Exception as e:
        return "unknown"
gc_utils._json_schema_to_python_type = patched_json_schema_to_python_type
# ====== END PATCHS ======

import spaces
import gradio as gr
from PIL import Image
import monkeypatch  # This file should be present to patch 
from src.tryon_pipeline import StableDiffusionXLInpaintPipeline as TryonPipeline
from src.unet_hacked_garmnet import UNet2DConditionModel as UNet2DConditionModel_ref
from src.unet_hacked_tryon import UNet2DConditionModel
from transformers import (
    CLIPImageProcessor,
    CLIPVisionModelWithProjection,
    CLIPTextModel,
    CLIPTextModelWithProjection,
)
from diffusers import DDPMScheduler, AutoencoderKL
from typing import List

import torch
import os
from transformers import AutoTokenizer

import numpy as np
from utils_mask import get_mask_location
from torchvision import transforms
import apply_net
from preprocess.humanparsing.run_parsing import Parsing
from preprocess.openpose.run_openpose import OpenPose
from detectron2.data.detection_utils import convert_PIL_to_numpy, _apply_exif_orientation
from torchvision.transforms.functional import to_pil_image

def pil_to_binary_mask(pil_image, threshold=0):
    np_image = np.array(pil_image)
    grayscale_image = Image.fromarray(np_image).convert("L")
    binary_mask = np.array(grayscale_image) > threshold
    mask = np.zeros(binary_mask.shape, dtype=np.uint8)
    for i in range(binary_mask.shape[0]):
        for j in range(binary_mask.shape[1]):
            if binary_mask[i, j]:
                mask[i, j] = 1
    mask = (mask * 255).astype(np.uint8)
    output_mask = Image.fromarray(mask)
    return output_mask

base_path = 'yisol/IDM-VTON'
example_path = os.path.join(os.path.dirname(__file__), 'example')

unet = UNet2DConditionModel.from_pretrained(
    base_path,
    subfolder="unet",
    torch_dtype=torch.float16,
)
unet.requires_grad_(False)

tokenizer_one = AutoTokenizer.from_pretrained(
    base_path,
    subfolder="tokenizer",
    revision=None,
    use_fast=False,
)
tokenizer_two = AutoTokenizer.from_pretrained(
    base_path,
    subfolder="tokenizer_2",
    revision=None,
    use_fast=False,
)
noise_scheduler = DDPMScheduler.from_pretrained(base_path, subfolder="scheduler")

text_encoder_one = CLIPTextModel.from_pretrained(
    base_path,
    subfolder="text_encoder",
    torch_dtype=torch.float16,
)
text_encoder_two = CLIPTextModelWithProjection.from_pretrained(
    base_path,
    subfolder="text_encoder_2",
    torch_dtype=torch.float16,
)
image_encoder = CLIPVisionModelWithProjection.from_pretrained(
    base_path,
    subfolder="image_encoder",
    torch_dtype=torch.float16,
)
vae = AutoencoderKL.from_pretrained(
    base_path,
    subfolder="vae",
    torch_dtype=torch.float16,
)

# "stabilityai/stable-diffusion-xl-base-1.0",
UNet_Encoder = UNet2DConditionModel_ref.from_pretrained(
    base_path,
    subfolder="unet_encoder",
    torch_dtype=torch.float16,
)

parsing_model = Parsing(0)
openpose_model = OpenPose(0)

UNet_Encoder.requires_grad_(False)
image_encoder.requires_grad_(False)
vae.requires_grad_(False)
unet.requires_grad_(False)
text_encoder_one.requires_grad_(False)
text_encoder_two.requires_grad_(False)

tensor_transfrom = transforms.Compose([
    transforms.ToTensor(),
    transforms.Normalize([0.5], [0.5]),
])

pipe = TryonPipeline.from_pretrained(
    base_path,
    unet=unet,
    vae=vae,
    feature_extractor=CLIPImageProcessor(),
    text_encoder=text_encoder_one,
    text_encoder_2=text_encoder_two,
    tokenizer=tokenizer_one,
    tokenizer_2=tokenizer_two,
    scheduler=noise_scheduler,
    image_encoder=image_encoder,
    torch_dtype=torch.float16,
)
pipe.unet_encoder = UNet_Encoder

@spaces.GPU
def start_tryon(dict, garm_img, garment_des, is_checked, is_checked_crop, denoise_steps, seed, category):
    """虚拟试衣主函数
    Args:
        dict: 输入图像字典，包含背景和图层信息
        garm_img: 服装图片
        garment_des: 服装描述文本
        is_checked: 是否启用自动检测模式
        is_checked_crop: 是否启用图像裁剪
        denoise_steps: 去噪步数
        seed: 随机种子
        category: 服装类别
    Returns:
        生成的试衣结果图像和灰度遮罩
    """
    device = "cuda"
    openpose_model.preprocessor.body_estimation.model.to(device)
    pipe.to(device)
    pipe.unet_encoder.to(device)

    # 2. 图像预处理 - 调整服装和人物图像大小
    garm_img = garm_img.convert("RGB").resize((768, 1024))
    human_img_orig = dict["background"].convert("RGB")
    orig_size = human_img_orig.size  # 保存原始尺寸

    # 2.1 如果启用裁剪，按3:4比例裁剪人物图像
    if is_checked_crop:
        width, height = human_img_orig.size
        target_width = int(min(width, height * (3 / 4)))
        target_height = int(min(height, width * (4 / 3)))
        left = (width - target_width) / 2
        top = (height - target_height) / 2
        right = (width + target_width) / 2
        bottom = (height + target_height) / 2
        cropped_img = human_img_orig.crop((left, top, right, bottom))
        crop_size = cropped_img.size
        human_img = cropped_img.resize((768, 1024))
    else:
        human_img = human_img_orig.resize((768, 1024))

    # 3. 生成遮罩
    if is_checked:
        # 3.1 使用自动检测模式
        keypoints = openpose_model(human_img.resize((384, 512)))
        model_parse, _ = parsing_model(human_img.resize((384, 512)))
        mask, mask_gray = get_mask_location('hd', category, model_parse, keypoints)
        mask = mask.resize((768, 1024))
    else:
        # 3.2 使用手动提供的遮罩
        mask = pil_to_binary_mask(dict['layers'][0].convert("RGB").resize((768, 1024)))

    # 3.3 生成灰度遮罩
    mask_gray = (1 - transforms.ToTensor()(mask)) * tensor_transfrom(human_img)
    mask_gray = to_pil_image((mask_gray + 1.0) / 2.0)

    # 4. 姿态处理
    human_img_arg = _apply_exif_orientation(human_img.resize((384, 512)))
    human_img_arg = convert_PIL_to_numpy(human_img_arg, format="BGR")
    args = apply_net.create_argument_parser().parse_args((
        'show',
        './configs/densepose_rcnn_R_50_FPN_s1x.yaml',
        './ckpt/densepose/model_final_162be9.pkl',
        'dp_segm',
        '-v',
        '--opts',
        'MODEL.DEVICE',
        'cuda'
    ))
    pose_img = args.func(args, human_img_arg)
    pose_img = pose_img[:, :, ::-1]
    pose_img = Image.fromarray(pose_img).resize((768, 1024))

    # 5. AI生成过程
    with torch.no_grad():
        with torch.cuda.amp.autocast():
            # 5.1 生成正面提示词嵌入
            prompt = "((best quality, masterpiece, ultra-detailed, high quality photography, photo realistic)), the model is wearing " + garment_des
            negative_prompt = "monochrome, lowres, bad anatomy, worst quality, normal quality, low quality, blurry, jpeg artifacts, sketch"
            with torch.inference_mode():
                (
                    prompt_embeds,
                    negative_prompt_embeds,
                    pooled_prompt_embeds,
                    negative_pooled_prompt_embeds,
                ) = pipe.encode_prompt(
                    prompt,
                    num_images_per_prompt=1,
                    do_classifier_free_guidance=True,
                    negative_prompt=negative_prompt,
                )

                # 5.2 生成服装相关的提示词嵌入
                prompt = "((best quality, masterpiece, ultra-detailed, high quality photography, photo realistic)), a photo of " + garment_des
                negative_prompt = "monochrome, lowres, bad anatomy, worst quality, normal quality, low quality, blurry, jpeg artifacts, sketch"
                if not isinstance(prompt, List):
                    prompt = [prompt] * 1
                if not isinstance(negative_prompt, List):
                    negative_prompt = [negative_prompt] * 1

                (
                    prompt_embeds_c,
                    _,
                    _,
                    _,
                ) = pipe.encode_prompt(
                    prompt,
                    num_images_per_prompt=1,
                    do_classifier_free_guidance=False,
                    negative_prompt=negative_prompt,
                )

                # 5.3 准备输入张量
                pose_tensor = tensor_transfrom(pose_img).unsqueeze(0).to(device, torch.float16)
                garm_tensor = tensor_transfrom(garm_img).unsqueeze(0).to(device, torch.float16)
                generator = torch.Generator(device).manual_seed(seed) if seed is not None else None

                # 6. 使用Stable Diffusion XL管道生成图像
                images = pipe(
                    prompt_embeds=prompt_embeds.to(device, torch.float16),
                    negative_prompt_embeds=negative_prompt_embeds.to(device, torch.float16),
                    pooled_prompt_embeds=pooled_prompt_embeds.to(device, torch.float16),
                    negative_pooled_prompt_embeds=negative_pooled_prompt_embeds.to(device, torch.float16),
                    num_inference_steps=denoise_steps,
                    generator=generator,
                    strength=1.0,
                    pose_img=pose_tensor.to(device, torch.float16),
                    text_embeds_cloth=prompt_embeds_c.to(device, torch.float16),
                    cloth=garm_tensor.to(device, torch.float16),
                    mask_image=mask,
                    image=human_img,
                    height=1024,
                    width=768,
                    ip_adapter_image=garm_img.resize((768, 1024)),
                    guidance_scale=2.0,
                )[0]

    # 7. 后处理 - 处理裁剪情况并返回结果
    if is_checked_crop:
        return images[0].resize(crop_size), mask_gray.resize(crop_size)
    else:
        return images[0].resize(orig_size), mask_gray.resize(orig_size)

# Setup example paths and lists
garm_list = os.listdir(os.path.join(example_path, "cloth"))
garm_list_path = [os.path.join(example_path, "cloth", garm) for garm in garm_list]

human_list = os.listdir(os.path.join(example_path, "human"))
human_list_path = [os.path.join(example_path, "human", human) for human in human_list]

human_ex_list = []
for ex_human in human_list_path:
    ex_dict = {}
    ex_dict['background'] = ex_human
    ex_dict['layers'] = None
    ex_dict['composite'] = None
    human_ex_list.append(ex_dict)

custom_css = """
:root {
    --primary: #9D4BFF;
    --secondary: #4A148C;
    --accent: #E0AAFF;
}

body {
    font-family: 'Helvetica Neue', sans-serif;
}

.purple-btn {
    background: var(--primary) !important;
    color: white !important;
    border: none !important;
    padding: 12px 24px !important;
    border-radius: 8px !important;
}

.purple-btn:hover {
    background: var(--secondary) !important;
}

.section-title {
    color: var(--secondary) !important;
    font-weight: 600 !important;
    margin-bottom: 10px !important;
}
"""

image_blocks = gr.Blocks(css=custom_css).queue()
with image_blocks as demo:
    gr.Markdown("## 👶 Baby Virtual Try-On Studio", elem_classes=["section-title"])
    
    # Coefficient Section (係數區塊)
    with gr.Column():
        try_button = gr.Button(
            value="✨ Generate Virtual Try-On", 
            elem_classes=["purple-btn"],
            scale=2
        )
        with gr.Accordion(label="Advanced Settings", open=False):
            with gr.Row():
                denoise_steps = gr.Number(label="Denoising Steps", minimum=20, maximum=40, value=30, step=1)
                seed = gr.Number(label="Seed", minimum=-1, maximum=2147483647, step=1, value=50)
    
    # Changing Section (更衣區塊)
    with gr.Row():
        with gr.Column():
            gr.Markdown("### 👶 Upload Baby Photo", elem_classes=["section-title"])
            imgs = gr.ImageEditor(sources='upload', type="pil", label='Human. Mask with pen or use auto-masking', interactive=True)
            with gr.Row():
                is_checked = gr.Checkbox(label="Yes", info="Use auto-generated mask (Takes 5 seconds)", value=True)
            with gr.Row():
                category = gr.Dropdown(
                    choices=["upper_body", "lower_body", "dresses"],
                    label="Category",
                    value="upper_body"
                )
            with gr.Row():
                is_checked_crop = gr.Checkbox(label="Yes", info="Use auto-crop & resizing", value=False)
            example = gr.Examples(
                inputs=imgs,
                examples_per_page=15,
                examples=human_ex_list
            )
    
        with gr.Column():
            gr.Markdown("### 👕 Upload Clothing", elem_classes=["section-title"])
            garm_img = gr.Image(label="Garment", sources='upload', type="pil")
            with gr.Row(elem_id="prompt-container"):
                prompt = gr.Textbox(label="Description of garment", placeholder="Short Sleeve Round Neck T-shirts", show_label=True, elem_id="prompt")
            example = gr.Examples(
                inputs=garm_img,
                examples_per_page=30,
                examples=garm_list_path
            )
    
        with gr.Column():
            image_out = gr.Image(label="Output", elem_id="output-img", show_share_button=False)
        with gr.Column():
            masked_img = gr.Image(label="Masked image output", elem_id="masked-img", show_share_button=False)
    
    with gr.Row():
        gr.Markdown("## Links")
    
    try_button.click(
        fn=start_tryon,
        inputs=[imgs, garm_img, prompt, is_checked, is_checked_crop, denoise_steps, seed, category],
        outputs=[image_out, masked_img],
        api_name='tryon'
    )

image_blocks.launch()