Hugging Face's logo

Spaces:
ryanlinjui
/
menu-text-detection
Running

App Files Community
menu-text-detection
File size: 4,883 Bytes 
8af6af2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
 
import os
import json

import numpy as np
import gradio as gr
from dotenv import load_dotenv

from menu.llm import (
    GeminiAPI,
    OpenAIAPI
)
from menu.donut import DonutFinetuned

load_dotenv()
GEMINI_API_TOKEN = os.getenv("GIMINI_API_TOKEN", "")
OPENAI_API_TOKEN = os.getenv("OPENAI_API_TOKEN", "")

SOURCE_CODE_GH_URL = "https://github.com/ryanlinjui/menu-text-detection"
BADGE_URL = "https://img.shields.io/badge/GitHub_Code-Click_Here!!-default?logo=github"

GITHUB_RAW_URL = "https://raw.githubusercontent.com/ryanlinjui/menu-text-detection/main"
EXAMPLE_IMAGE_LIST = [
    f"{GITHUB_RAW_URL}/examples/menu-hd.jpg",
    f"{GITHUB_RAW_URL}/examples/menu-vs.jpg",
    f"{GITHUB_RAW_URL}/examples/menu-si.jpg"
]
MODEL_LIST = [
    "Donut Model",
    "gemini-2.0-flash",
    "gemini-2.5-flash-preview-04-17",
    "gemini-2.5-pro-preview-03-25",
    "gpt-4.1",
    "gpt-4o",
    "o4-mini"
]

def handle(image: np.ndarray, model: str, api_token: str) -> str:
    if image is None:
        raise gr.Error("Please upload an image first.")
    
    if model == MODEL_LIST[0]:
        result = DonutFinetuned.predict(image)
    
    elif model in MODEL_LIST[1:]:
        if len(api_token) < 10:
            raise gr.Error(f"Please provide a valid token for {model}.")
        try:
            if model in MODEL_LIST[1:4]:
                result = GeminiAPI.call(image, model, api_token)
            else:
                result = OpenAIAPI.call(image, model, api_token)
        except Exception as e:
            raise gr.Error(f"Failed to process with API model {model}: {str(e)}")
    else:
        raise gr.Error("Invalid model selection. Please choose a valid model.")
    
    return json.dumps(result, indent=4, ensure_ascii=False)

def UserInterface() -> gr.Interface:
    with gr.Blocks(
        delete_cache=(86400, 86400),
        css="""
        .image-panel {
            display: flex;
            flex-direction: column;
            height: 600px;
        }
        .image-panel img {
            object-fit: contain;
            max-height: 600px;
            max-width: 600px;
            width: 100%;
        }
        .large-text textarea {
            font-size: 20px !important;
            height: 600px !important;
            width: 100% !important;
        }
        """
    ) as gradio_interface:
        gr.HTML(f'<a href="{SOURCE_CODE_GH_URL}"><img src="{BADGE_URL}" alt="GitHub Code"/></a>')
        gr.Markdown("# Menu Text Detection")
        
        with gr.Row():
            with gr.Column(scale=1, min_width=500):
                gr.Markdown("## 📷 Menu Image")
                menu_image = gr.Image(
                    type="numpy", 
                    label="Input menu image",
                    elem_classes="image-panel"
                )
                
                gr.Markdown("## 🤖 Model Selection")
                model_choice_dropdown = gr.Dropdown(
                    choices=MODEL_LIST,
                    value=MODEL_LIST[0],
                    label="Select Text Detection Model"
                )
                
                api_token_textbox = gr.Textbox(
                    label="API Token",
                    placeholder="Enter your API token here...",
                    type="password",
                    visible=False
                )
                
                generate_button = gr.Button("Generate Menu Information", variant="primary")

                gr.Examples(
                    examples=EXAMPLE_IMAGE_LIST,
                    inputs=menu_image,
                    label="Example Menu Images"
                )
            
            with gr.Column(scale=1):
                gr.Markdown("## 🍽️ Menu Info")
                menu_json_textbox = gr.Textbox(
                    label="Ouput JSON",
                    interactive=False,
                    text_align="left",
                    elem_classes="large-text"
                )
        
        def update_token_visibility(choice):
            if choice in MODEL_LIST[1:]:
                current_token = ""
                if choice in MODEL_LIST[1:4]:
                    current_token = GEMINI_API_TOKEN
                elif choice in MODEL_LIST[4:]:
                    current_token = OPENAI_API_TOKEN
                return gr.update(visible=True, value=current_token)
            else:
                return gr.update(visible=False)
                
        model_choice_dropdown.change(
            fn=update_token_visibility,
            inputs=model_choice_dropdown,
            outputs=api_token_textbox
        )
        
        generate_button.click(
            fn=handle,
            inputs=[menu_image, model_choice_dropdown, api_token_textbox],
            outputs=menu_json_textbox
        )

    return gradio_interface

if __name__ == "__main__":
    demo = UserInterface()
    demo.launch()