File size: 4,838 Bytes
5b344d3
 
 
 
 
 
ec86d75
 
5b344d3
f68c7aa
5b344d3
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3c0e164
56db7c1
5b344d3
 
ec86d75
 
 
b44d446
 
ec86d75
 
 
d87c251
ec86d75
5b344d3
 
 
 
 
 
 
 
 
 
 
de651c2
5b344d3
de651c2
5b344d3
 
 
 
 
 
 
 
 
 
 
de651c2
5b344d3
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
from typing import Optional

import numpy as np
import cv2
import streamlit as st
from PIL import Image
import os
import tempfile

from sdfile import PIPELINES, generate

DEFAULT_PROMPT = "belted shirt black belted portrait-collar wrap blouse with black prints"
DEAFULT_WIDTH, DEFAULT_HEIGHT = 512,512
OUTPUT_IMAGE_KEY = "output_img"
LOADED_IMAGE_KEY = "loaded_img"

def get_image(key: str) -> Optional[Image.Image]:
    if key in st.session_state:
        return st.session_state[key]
    return None

def set_image(key:str, img: Image.Image):
    st.session_state[key] = img

def prompt_and_generate_button(prefix, pipeline_name: PIPELINES, **kwargs):
    prompt = st.text_area(
        "Prompt",
        value = DEFAULT_PROMPT,
        key = f"{prefix}-prompt"
    )
    negative_prompt = st.text_area(
        "Negative prompt",
        value = "",
        key =f"{prefix}-negative_prompt",
    )
    col1,col2 =st.columns(2)
    with col1:
        steps = st.slider(
            "Number of inference steps",
            min_value=1,
            max_value=200,
            value=30,
            key=f"{prefix}-inference-steps",
        )
    with col2:
        guidance_scale = st.slider(
            "Guidance scale",
            min_value=0.0,
            max_value=20.0,
            value= 7.5,
            step = 0.5,
            key=f"{prefix}-guidance-scale",
        )
    enable_cpu_offload = st.checkbox(
        "Enable CPU offload if you run out of memory",
        key =f"{prefix}-cpu-offload",
        value= False,
    )

    if st.button("Generate Image", key = f"{prefix}-btn"):
        with st.spinner("Generating image ..."):
            image = generate(
                prompt,
                pipeline_name,
                negative_prompt=negative_prompt,
                num_inference_steps=steps,
                guidance_scale=guidance_scale,
                enable_cpu_offload=enable_cpu_offload,
                **kwargs,
            )
            set_image(OUTPUT_IMAGE_KEY,image.copy())
        st.image(image)
def width_and_height_sliders(prefix):
    col1, col2 = st.columns(2)
    with col1:
        width = st.slider(
            "Width",
            min_value=64,
            max_value=1600,
            step=16,
            value=512,
            key=f"{prefix}-width",
        )
    with col2:
        height = st.slider(
            "Height",
            min_value=64,
            max_value=1600,
            step=16,
            value=512,
            key=f"{prefix}-height",
        )
    return width, height

def image_uploader(prefix):
    image = st.file_uploader("Image", ["jpg", "png"], key=f"{prefix}-uploader")
    if image:
        image = Image.open(image)
        print(f"loaded input image of size ({image.width}, {image.height})")
        return image

    return get_image(LOADED_IMAGE_KEY)

def sketching():
    image = image_uploader("sketch2img")
    
    if not image:
        return None,None
        
    with tempfile.TemporaryDirectory() as temp_dir:
        temp_image_path = os.path.join(temp_dir, "uploaded_image.jpg")
        image.save(temp_image_path)
        
        image = cv2.imread(temp_image_path)
        image = cv2.cvtColor(image,cv2.COLOR_BGR2GRAY)
        image_blur = cv2.GaussianBlur(image,(5,5),0)
        sketch = cv2.adaptiveThreshold(image_blur, 255, cv2.ADAPTIVE_THRESH_MEAN_C,cv2.THRESH_BINARY,11,2)
        sketch_pil = Image.fromarray(sketch)
    return sketch_pil

def txt2img_tab():
    prefix = "txt2img"
    width, height = width_and_height_sliders(prefix)
    prompt_and_generate_button(prefix,"txt2img",width=width,height=height)

def sketching_tab():
    prefix = "sketch2img"
    col1,col2 = st.columns(2)
    with col1:
        image = sketching()
    with col2:
        if image:
            controlnet_conditioning_scale = st.slider(
                "Strength or dependence on the input sketch",
                min_value=0.0,
                max_value= 1.0,
                value = 0.5,
                step = 0.05,
                key=f"{prefix}-controlnet_conditioning_scale",
            )
            prompt_and_generate_button(
                prefix,
                "sketch2img",
                image=image,
                controlnet_conditioning_scale=controlnet_conditioning_scale,
            )

def main():
    st.set_page_config(layout="wide")
    st.title("Fashion-SDX: Playground")

    tab1,tab2 = st.tabs(
        ["Text to image", "Sketch to image"]
    )
    with tab1:
        txt2img_tab()
    with tab2:
        sketching_tab()

    with st.sidebar:
        st.header("Most Recent Output Image")
        output_image = get_image((OUTPUT_IMAGE_KEY))
        if output_image:
            st.image(output_image)
        else:
            st.markdown("no output generated yet")
if __name__ =="__main__":
    main()