Spaces:
Sleeping
Sleeping
| import sys | |
| sys.path.append('./') | |
| from PIL import Image | |
| try: | |
| import cv2 | |
| print("OpenCV is installed correctly.") | |
| except ImportError: | |
| print("OpenCV is not installed.") | |
| import gradio as gr | |
| 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 spaces | |
| 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] == True : | |
| 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 | |
| # Function to visualize parsing | |
| def visualize_parsing(image, mask): | |
| """ | |
| Visualize the parsing by applying a color map to the segmentation mask. | |
| """ | |
| # Ensure image is in RGB format and convert to numpy array | |
| image_array = np.array(image.convert('RGB'), dtype=np.uint8) | |
| # Create a color map | |
| num_classes = np.max(mask) + 1 | |
| colors = np.random.randint(0, 255, size=(num_classes, 3), dtype=np.uint8) | |
| # Apply color map to the mask | |
| color_mask = colors[mask.astype(int)] | |
| # Ensure color_mask is correctly shaped and typed | |
| color_mask = np.array(color_mask, dtype=np.uint8) | |
| # Combine the original image and the color mask | |
| combined_image = cv2.addWeighted(image_array, 0.5, color_mask, 0.5, 0) | |
| return Image.fromarray(combined_image) | |
| def process_densepose(human_img): | |
| """ | |
| Processes the human image using DensePose and returns the DensePose image. | |
| Assumes human_img is a dictionary with a 'background' key pointing to the image path. | |
| """ | |
| # Load image from path | |
| device = "cuda" | |
| image_path = human_img['background'] # Assuming 'background' is the correct key | |
| if isinstance(image_path, Image.Image): | |
| image = image_path | |
| else: | |
| image = Image.open(image_path) # Only call Image.open if it's not already an Image object | |
| # Apply EXIF orientation and resize | |
| human_img_arg = _apply_exif_orientation(image.resize((384, 512))) | |
| human_img_arg = convert_PIL_to_numpy(human_img_arg, format="BGR") | |
| # Setup DensePose arguments | |
| 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] # Convert from BGR to RGB | |
| pose_img = Image.fromarray(pose_img).resize((768, 1024)) | |
| return pose_img, pose_img | |
| def process_human_parsing(human_img): | |
| """ | |
| Processes the human image to perform segmentation using a human parsing model. | |
| """ | |
| image_path = human_img['background'] # Assuming 'background' is the correct key | |
| if isinstance(image_path, Image.Image): | |
| image = image_path | |
| else: | |
| image = Image.open(image_path) # Only call Image.open if it's not already an Image object | |
| image = image.resize((384, 512)) | |
| model_parse, _ = parsing_model(image) | |
| # parsing_image = visualize_parsing(human_img, model_parse) # Visualization function needed | |
| # vis_image = visualize_parsing(image, model_parse) | |
| # state_message = "Human parsing processing completed" | |
| return model_parse | |
| def start_tryon(dict, garm_img, garment_des, is_checked, is_checked_crop, denoise_steps, seed): | |
| """ | |
| Preprocesses images and generates outputs using various models. | |
| Parameters: | |
| - human_img: PIL image of the human. | |
| - garm_img: PIL image of the garment. | |
| - garment_des: Description of the garment. | |
| - is_checked: Boolean flag indicating whether to use auto-generated mask. | |
| - is_checked_crop: Boolean flag indicating whether to use auto-crop & resizing. | |
| - denoise_steps: Number of denoising steps. | |
| - seed: Seed for random generator. | |
| - pose_img: DensePose image generated in the previous step. | |
| Returns: | |
| - Processed images: Depending on the conditions, it returns human_img_orig, mask_gray, and final output images. | |
| """ | |
| openpose_model.preprocessor.body_estimation.model.to(device) | |
| pipe.to(device) | |
| pipe.unet_encoder.to(device) | |
| garm_img= garm_img.convert("RGB").resize((768,1024)) | |
| human_img_orig = dict["background"].convert("RGB") | |
| 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)) | |
| if is_checked: | |
| keypoints = openpose_model(human_img.resize((384,512))) | |
| print(keypoints) | |
| model_parse, _ = parsing_model(human_img.resize((384,512))) | |
| print(model_parse) | |
| mask, mask_gray = get_mask_location('hd', "upper_body", model_parse, keypoints) | |
| mask = mask.resize((768,1024)) | |
| else: | |
| mask = pil_to_binary_mask(dict['layers'][0].convert("RGB").resize((768, 1024))) | |
| # mask = transforms.ToTensor()(mask) | |
| # mask = mask.unsqueeze(0) | |
| mask_gray = (1-transforms.ToTensor()(mask)) * tensor_transfrom(human_img) | |
| mask_gray = to_pil_image((mask_gray+1.0)/2.0) | |
| 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')) | |
| # verbosity = getattr(args, "verbosity", None) | |
| pose_img = args.func(args,human_img_arg) | |
| pose_img = pose_img[:,:,::-1] | |
| pose_img = Image.fromarray(pose_img).resize((768,1024)) | |
| with torch.no_grad(): | |
| # Extract the images | |
| with torch.cuda.amp.autocast(): | |
| with torch.no_grad(): | |
| prompt = "model is wearing " + garment_des | |
| negative_prompt = "monochrome, lowres, bad anatomy, worst quality, low quality" | |
| 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, | |
| ) | |
| prompt = "a photo of " + garment_des | |
| negative_prompt = "monochrome, lowres, bad anatomy, worst quality, low quality" | |
| if not isinstance(prompt, List): | |
| prompt = [prompt] * 1 | |
| if not isinstance(negative_prompt, List): | |
| negative_prompt = [negative_prompt] * 1 | |
| with torch.inference_mode(): | |
| ( | |
| prompt_embeds_c, | |
| _, | |
| _, | |
| _, | |
| ) = pipe.encode_prompt( | |
| prompt, | |
| num_images_per_prompt=1, | |
| do_classifier_free_guidance=False, | |
| negative_prompt=negative_prompt, | |
| ) | |
| pose_img = 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 | |
| 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_img.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] | |
| if is_checked_crop: | |
| out_img = images[0].resize(crop_size) | |
| human_img_orig.paste(out_img, (int(left), int(top))) | |
| return human_img_orig, mask_gray | |
| else: | |
| # out_img = images[0].resize(crop_size) | |
| return images[0], mask_gray | |
| 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) | |
| ##default human | |
| image_blocks = gr.Blocks().queue() | |
| with image_blocks as demo: | |
| with gr.Row(): | |
| with gr.Column(): | |
| 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(): | |
| is_checked_crop = gr.Checkbox(label="Yes", info="Use auto-crop & resizing",value=False) | |
| example = gr.Examples( | |
| inputs=imgs, | |
| examples_per_page=10, | |
| examples=human_ex_list | |
| ) | |
| with gr.Column(): | |
| garm_img = gr.Image(label="Garment", sources='upload', type="pil") | |
| with gr.Row(elem_id="prompt-container"): | |
| with gr.Row(): | |
| prompt = gr.Textbox(placeholder="Description of garment ex) Short Sleeve Round Neck T-shirts", show_label=False, elem_id="prompt") | |
| example = gr.Examples( | |
| inputs=garm_img, | |
| examples_per_page=8, | |
| examples=garm_list_path) | |
| with gr.Column(): | |
| # image_out = gr.Image(label="Output", elem_id="output-img", height=400) | |
| masked_img = gr.Image(label="Masked image output", elem_id="masked-img",show_share_button=False) | |
| with gr.Column(): | |
| # image_out = gr.Image(label="Output", elem_id="output-img", height=400) | |
| image_out = gr.Image(label="Output", elem_id="output-img",show_share_button=False) | |
| with gr.Column(): | |
| densepose_img_out = gr.Image(label="Output", elem_id="densepose-img",show_share_button=False) | |
| # densepose_img = gr.Gallery(label="All images", show_label=False, elem_id="all-images", columns=[3], rows=[1], object_fit="contain", height="auto") | |
| with gr.Column(): | |
| try_button = gr.Button(value="Try-on") | |
| 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=42) | |
| densepose_state = gr.State(None) | |
| # Define the steps in sequence | |
| image_blocks = gr.Blocks().queue() | |
| with image_blocks as demo: | |
| with gr.Row(): | |
| with gr.Column(): | |
| 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(): | |
| is_checked_crop = gr.Checkbox(label="Yes", info="Use auto-crop & resizing",value=False) | |
| example = gr.Examples( | |
| inputs=imgs, | |
| examples_per_page=10, | |
| examples=human_ex_list | |
| ) | |
| with gr.Column(): | |
| garm_img = gr.Image(label="Garment", sources='upload', type="pil") | |
| with gr.Row(elem_id="prompt-container"): | |
| with gr.Row(): | |
| prompt = gr.Textbox(placeholder="Description of garment ex) Short Sleeve Round Neck T-shirts", show_label=False, elem_id="prompt") | |
| example = gr.Examples( | |
| inputs=garm_img, | |
| examples_per_page=8, | |
| examples=garm_list_path) | |
| with gr.Column(): | |
| masked_img = gr.Image(label="Masked image output", elem_id="masked-img", show_share_button=False) | |
| with gr.Column(): | |
| image_out = gr.Image(label="Output", elem_id="output-img", show_share_button=False) | |
| with gr.Column(): | |
| densepose_img_out = gr.Image(label="Dense-pose", elem_id="densepose-img", show_share_button=False) | |
| # densepose_img = gr.Gallery(label="All images", show_label=False, elem_id="all-images", columns=[3], rows=[1], object_fit="contain", height="auto") | |
| with gr.Column(): | |
| human_parse_img_out = gr.Image(label="Human-Parse", elem_id="humanparse-img", show_share_button=False) | |
| # densepose_img = gr.Gallery(label="All images", show_label=False, elem_id="all-images", columns=[3], rows=[1], object_fit="contain", height="auto") | |
| with gr.Column(): | |
| try_button = gr.Button(value="Try-on") | |
| get_denspose =gr.Button(value="Get-DensePose") | |
| get_humanparse =gr.Button(value="Get-HumanParse") | |
| 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=42) | |
| densepose_state = gr.State(None) | |
| # Define the steps in sequence | |
| get_denspose.click( | |
| fn=process_densepose, | |
| inputs=[imgs], | |
| outputs=[densepose_img_out, densepose_state], | |
| api_name='process_densepose' | |
| ) | |
| get_humanparse.click( | |
| fn=process_human_parsing, | |
| inputs=[imgs], | |
| outputs=[human_parse_img_out], | |
| api_name='process_humanparse' | |
| ) | |
| try_button.click( | |
| fn=start_tryon, | |
| inputs=[imgs, garm_img, prompt, is_checked, is_checked_crop, denoise_steps, seed], | |
| outputs=[image_out, masked_img], | |
| api_name='start_tryon' | |
| ) | |
| image_blocks.launch(server_name="0.0.0.0", server_port=3000) | |