File size: 4,066 Bytes
c06e0c5 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 |
---
frameworks:
- Pytorch
license: Apache License 2.0
tags: []
tasks:
- image-to-image
base_model:
- Qwen/Qwen-Image-Edit
base_model_relation: adapter
---
# Qwen-Image-Edit 人脸生成图像模型
## 模型介绍
本模型是基于 [Qwen-Image-Edit](https://www.modelscope.cn/models/Qwen/Qwen-Image-Edit) 人脸控制图像生成模型。输入裁剪下的人脸图像，输出该人的人像图片。
## 效果展示
|人脸|生成图1|生成图2|生成图3|生成图4|
|-|-|-|-|-|
||||||
## 推理代码
```
git clone https://github.com/modelscope/DiffSynth-Studio.git
cd DiffSynth-Studio
pip install -e .
```
```python
from diffsynth.pipelines.qwen_image import QwenImagePipeline, ModelConfig
import torch
from modelscope import snapshot_download, dataset_snapshot_download
from PIL import Image
pipe = QwenImagePipeline.from_pretrained(
torch_dtype=torch.bfloat16,
device="cuda",
model_configs=[
ModelConfig(model_id="Qwen/Qwen-Image-Edit", origin_file_pattern="transformer/diffusion_pytorch_model*.safetensors"),
ModelConfig(model_id="Qwen/Qwen-Image", origin_file_pattern="text_encoder/model*.safetensors"),
ModelConfig(model_id="Qwen/Qwen-Image", origin_file_pattern="vae/diffusion_pytorch_model.safetensors"),
],
tokenizer_config=None,
processor_config=ModelConfig(model_id="Qwen/Qwen-Image-Edit", origin_file_pattern="processor/"),
)
snapshot_download("DiffSynth-Studio/Qwen-Image-Edit-F2P", local_dir="models/DiffSynth-Studio/Qwen-Image-Edit-F2P", allow_file_pattern="model.safetensors")
pipe.load_lora(pipe.dit, "models/DiffSynth-Studio/Qwen-Image-Edit-F2P/model.safetensors")
dataset_snapshot_download(
dataset_id="DiffSynth-Studio/example_image_dataset",
local_dir="./data/example_image_dataset",
allow_file_pattern="f2p/qwen_woman_face_crop.png"
)
face_image = Image.open("data/example_image_dataset/f2p/qwen_woman_face_crop.png").convert("RGB")
prompt = "摄影。一个年轻女性穿着黄色连衣裙，站在花田中，背景是五颜六色的花朵和绿色的草地。"
image = pipe(prompt, edit_image=face_image, seed=42, num_inference_steps=40, height=1152, width=864)
image.save(f"image.jpg")
```
人脸自动裁剪
```python
import torch
from PIL import Image
import numpy as np
from insightface.app import FaceAnalysis
import cv2
class FaceDetector(torch.nn.Module):
def __init__(self):
super().__init__()
providers = ["CUDAExecutionProvider", "CPUExecutionProvider"]
provider_options = [{"device_id": 0}, {}]
self.app_640 = FaceAnalysis(name='antelopev2', providers=providers, provider_options=provider_options)
self.app_640.prepare(ctx_id=0, det_size=(640, 640))
self.app_320 = FaceAnalysis(name='antelopev2', providers=providers, provider_options=provider_options)
self.app_320.prepare(ctx_id=0, det_size=(320, 320))
self.app_160 = FaceAnalysis(name='antelopev2', providers=providers, provider_options=provider_options)
self.app_160.prepare(ctx_id=0, det_size=(160, 160))
def _detect_face(self, id_image_cv2):
face_info = self.app_640.get(id_image_cv2)
if len(face_info) > 0:
return face_info
face_info = self.app_320.get(id_image_cv2)
if len(face_info) > 0:
return face_info
face_info = self.app_160.get(id_image_cv2)
return face_info
def crop_face(self, id_image):
face_info = self._detect_face(cv2.cvtColor(np.array(id_image), cv2.COLOR_RGB2BGR))
if len(face_info) == 0:
return None
else:
bbox = sorted(face_info, key=lambda x: (x['bbox'][2] - x['bbox'][0]) * (x['bbox'][3] - x['bbox'][1]))[-1]['bbox']
return id_image.crop(list(map(int, bbox)))
face_detector = FaceDetector()
face_image = face_detector.crop_face(Image.open("image_2.jpg"))
face_image.save("face_crop.jpg")
``` |