app.py

import os
import numpy as np
import torch
import torch.nn as nn
import gradio as gr
from torchvision.models import efficientnet_v2_m, EfficientNet_V2_M_Weights
import torch.nn.functional as F
from torchvision import transforms
from PIL import Image
from data_manager import get_dog_description
from urllib.parse import quote
os.system('pip install ultralytics')
from ultralytics import YOLO

# 下載YOLOv5預訓練模型
model_yolo = YOLO('yolov5s.pt')  # 使用 YOLOv5 預訓練模型


dog_breeds = ["Afghan_Hound", "African_Hunting_Dog", "Airedale", "American_Staffordshire_Terrier", 
              "Appenzeller", "Australian_Terrier", "Bedlington_Terrier", "Bernese_Mountain_Dog", 
              "Blenheim_Spaniel", "Border_Collie", "Border_Terrier", "Boston_Bull", "Bouvier_Des_Flandres", 
              "Brabancon_Griffon", "Brittany_Spaniel", "Cardigan", "Chesapeake_Bay_Retriever", 
              "Chihuahua", "Dandie_Dinmont", "Doberman", "English_Foxhound", "English_Setter", 
              "English_Springer", "EntleBucher", "Eskimo_Dog", "French_Bulldog", "German_Shepherd", 
              "German_Short-Haired_Pointer", "Gordon_Setter", "Great_Dane", "Great_Pyrenees", 
              "Greater_Swiss_Mountain_Dog", "Ibizan_Hound", "Irish_Setter", "Irish_Terrier", 
              "Irish_Water_Spaniel", "Irish_Wolfhound", "Italian_Greyhound", "Japanese_Spaniel", 
              "Kerry_Blue_Terrier", "Labrador_Retriever", "Lakeland_Terrier", "Leonberg", "Lhasa", 
              "Maltese_Dog", "Mexican_Hairless", "Newfoundland", "Norfolk_Terrier", "Norwegian_Elkhound", 
              "Norwich_Terrier", "Old_English_Sheepdog", "Pekinese", "Pembroke", "Pomeranian", 
              "Rhodesian_Ridgeback", "Rottweiler", "Saint_Bernard", "Saluki", "Samoyed", 
              "Scotch_Terrier", "Scottish_Deerhound", "Sealyham_Terrier", "Shetland_Sheepdog", 
              "Shih-Tzu", "Siberian_Husky", "Staffordshire_Bullterrier", "Sussex_Spaniel", 
              "Tibetan_Mastiff", "Tibetan_Terrier", "Walker_Hound", "Weimaraner", 
              "Welsh_Springer_Spaniel", "West_Highland_White_Terrier", "Yorkshire_Terrier", 
              "Affenpinscher", "Basenji", "Basset", "Beagle", "Black-and-Tan_Coonhound", "Bloodhound", 
              "Bluetick", "Borzoi", "Boxer", "Briard", "Bull_Mastiff", "Cairn", "Chow", "Clumber", 
              "Cocker_Spaniel", "Collie", "Curly-Coated_Retriever", "Dhole", "Dingo", 
              "Flat-Coated_Retriever", "Giant_Schnauzer", "Golden_Retriever", "Groenendael", "Keeshond", 
              "Kelpie", "Komondor", "Kuvasz", "Malamute", "Malinois", "Miniature_Pinscher", 
              "Miniature_Poodle", "Miniature_Schnauzer", "Otterhound", "Papillon", "Pug", "Redbone", 
              "Schipperke", "Silky_Terrier", "Soft-Coated_Wheaten_Terrier", "Standard_Poodle", 
              "Standard_Schnauzer", "Toy_Poodle", "Toy_Terrier", "Vizsla", "Whippet", 
              "Wire-Haired_Fox_Terrier"]

class MultiHeadAttention(nn.Module):

    def __init__(self, in_dim, num_heads=8):
        super().__init__()
        self.num_heads = num_heads
        self.head_dim = max(1, in_dim // num_heads)
        self.scaled_dim = self.head_dim * num_heads
        self.fc_in = nn.Linear(in_dim, self.scaled_dim)
        self.query = nn.Linear(self.scaled_dim, self.scaled_dim)
        self.key = nn.Linear(self.scaled_dim, self.scaled_dim)
        self.value = nn.Linear(self.scaled_dim, self.scaled_dim)
        self.fc_out = nn.Linear(self.scaled_dim, in_dim)

    def forward(self, x):
        N = x.shape[0]
        x = self.fc_in(x)
        q = self.query(x).view(N, self.num_heads, self.head_dim)
        k = self.key(x).view(N, self.num_heads, self.head_dim)
        v = self.value(x).view(N, self.num_heads, self.head_dim)

        energy = torch.einsum("nqd,nkd->nqk", [q, k])
        attention = F.softmax(energy / (self.head_dim ** 0.5), dim=2)

        out = torch.einsum("nqk,nvd->nqd", [attention, v])
        out = out.reshape(N, self.scaled_dim)
        out = self.fc_out(out)
        return out

class BaseModel(nn.Module):
    def __init__(self, num_classes, device='cuda' if torch.cuda.is_available() else 'cpu'):
        super().__init__()
        self.device = device
        self.backbone = efficientnet_v2_m(weights=EfficientNet_V2_M_Weights.IMAGENET1K_V1)
        self.feature_dim = self.backbone.classifier[1].in_features
        self.backbone.classifier = nn.Identity()

        self.num_heads = max(1, min(8, self.feature_dim // 64))
        self.attention = MultiHeadAttention(self.feature_dim, num_heads=self.num_heads)

        self.classifier = nn.Sequential(
            nn.LayerNorm(self.feature_dim),
            nn.Dropout(0.3),
            nn.Linear(self.feature_dim, num_classes)
        )

        self.to(device)

    def forward(self, x):
        x = x.to(self.device)
        features = self.backbone(x)
        attended_features = self.attention(features)
        logits = self.classifier(attended_features)
        return logits, attended_features


num_classes = 120
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
model = BaseModel(num_classes=num_classes, device=device)

checkpoint = torch.load('best_model_81_dog.pth', map_location=torch.device('cpu'))
model.load_state_dict(checkpoint['model_state_dict'])

# evaluation mode
model.eval()

# Image preprocessing function
def preprocess_image(image):
    # If the image is numpy.ndarray turn into PIL.Image
    if isinstance(image, np.ndarray):
        image = Image.fromarray(image)

    # Use torchvision.transforms to process images
    transform = transforms.Compose([
        transforms.Resize((224, 224)),
        transforms.ToTensor(),
        transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]),
    ])

    return transform(image).unsqueeze(0)


def get_akc_breeds_link():
    return "https://www.akc.org/dog-breeds/"

# def predict(image):
#     if image is None:
#         return "Please upload an image to get started.", gr.update(visible=False), gr.update(visible=False), gr.update(visible=False)
    
#     try:
#         image_tensor = preprocess_image(image)
#         with torch.no_grad():
#             output = model(image_tensor)
#             logits = output[0] if isinstance(output, tuple) else output

#             probabilities = F.softmax(logits, dim=1)
#             topk_probs, topk_indices = torch.topk(probabilities, k=3)

#             top1_prob = topk_probs[0][0].item()
#             topk_breeds = [dog_breeds[idx.item()] for idx in topk_indices[0]]
#             topk_probs_percent = [f"{prob.item() * 100:.2f}%" for prob in topk_probs[0]]

#             if top1_prob >= 0.5:
#                 breed = topk_breeds[0]
#                 description = get_dog_description(breed)
#                 return format_description(description, breed), gr.update(visible=False), gr.update(visible=False), gr.update(visible=False)

#             elif top1_prob < 0.2:
#                 return ("The image is too unclear or the dog breed is not in the dataset. Please upload a clearer image of the dog.",
#                         gr.update(visible=False), gr.update(visible=False), gr.update(visible=False))
#             else:
#                 explanation = (
#                     f"The model couldn't confidently identify the breed. Here are the top 3 possible breeds:\n\n"
#                     f"1. **{topk_breeds[0]}** ({topk_probs_percent[0]} confidence)\n"
#                     f"2. **{topk_breeds[1]}** ({topk_probs_percent[1]} confidence)\n"
#                     f"3. **{topk_breeds[2]}** ({topk_probs_percent[2]} confidence)\n\n"
#                     "Click on a button to view more information about the breed."
#                 )
#                 return explanation, gr.update(visible=True, value=f"More about {topk_breeds[0]}"), gr.update(visible=True, value=f"More about {topk_breeds[1]}"), gr.update(visible=True, value=f"More about {topk_breeds[2]}")

#     except Exception as e:
#         return f"An error occurred: {e}", gr.update(visible=False), gr.update(visible=False), gr.update(visible=False)


def predict(image):
    if image is None:
        return "Please upload an image to get started.", gr.update(visible=False), gr.update(visible=False), gr.update(visible=False)

    try:
        # 檢查圖片是否是 numpy.ndarray，如果是則轉換為 PIL.Image
        if isinstance(image, np.ndarray):
            image = Image.fromarray(image)

        # 使用 YOLO 偵測狗
        results = model_yolo(image)
        boxes = results[0].boxes  # 提取邊界框

        if len(boxes) == 0:
            return "No dog detected in the image.", gr.update(visible=False), gr.update(visible=False), gr.update(visible=False)

        explanations = []
        visible_buttons = []

        for i, box in enumerate(boxes):
            # 提取每隻狗的區域
            x1, y1, x2, y2 = map(int, box.xyxy[0])  # 使用 box.xyxy 來提取邊界框座標

            # 裁剪出狗區域，確保 image 是 PIL.Image 格式
            cropped_image = image.crop((x1, y1, x2, y2))
            image_tensor = preprocess_image(cropped_image)

            with torch.no_grad():
                output = model(image_tensor)
                logits = output[0] if isinstance(output, tuple) else output

                probabilities = F.softmax(logits, dim=1)
                topk_probs, topk_indices = torch.topk(probabilities, k=3)

                top1_prob = topk_probs[0][0].item()
                topk_breeds = [dog_breeds[idx.item()] for idx in topk_indices[0]]
                topk_probs_percent = [f"{prob.item() * 100:.2f}%" for prob in topk_probs[0]]

                # 根據信心分數進行判斷
                if top1_prob >= 0.5:
                    breed = topk_breeds[0]
                    description = get_dog_description(breed)
                    explanations.append(f"Detected a dog: **{breed}** with {topk_probs_percent[0]} confidence.")
                elif 0.2 <= top1_prob < 0.5:
                    explanation = (
                        f"Detected a dog with moderate confidence. Here are the top 3 possible breeds:\n"
                        f"1. **{topk_breeds[0]}** ({topk_probs_percent[0]} confidence)\n"
                        f"2. **{topk_breeds[1]}** ({topk_probs_percent[1]} confidence)\n"
                        f"3. **{topk_breeds[2]}** ({topk_probs_percent[2]} confidence)\n"
                    )
                    explanations.append(explanation)
                    visible_buttons.extend([i+1 for _ in range(3)])
                else:
                    explanations.append("The image is too unclear or the breed is not in the dataset. Please upload a clearer image.")

        # 處理不同情境的結果
        if len(explanations) > 0:
            final_explanation = "\n\n".join(explanations)
            return final_explanation, gr.update(visible=len(visible_buttons) >= 1), gr.update(visible=len(visible_buttons) >= 2), gr.update(visible=len(visible_buttons) >= 3)

    except Exception as e:
        return f"An error occurred: {e}", gr.update(visible=False), gr.update(visible=False), gr.update(visible=False)




def format_description(description, breed):
    if isinstance(description, dict):
        formatted_description = "\n\n".join([f"**{key}**: {value}" for key, value in description.items()])
    else:
        formatted_description = description

    akc_link = get_akc_breeds_link()
    formatted_description += f"\n\n**Want to learn more about dog breeds?** [Visit the AKC dog breeds page]({akc_link}) and search for {breed} to find detailed information."

    disclaimer = ("\n\n*Disclaimer: The external link provided leads to the American Kennel Club (AKC) dog breeds page. "
                  "You may need to search for the specific breed on that page. "
                  "I am not responsible for the content on external sites. "
                  "Please refer to the AKC's terms of use and privacy policy.*")
    formatted_description += disclaimer

    return formatted_description

def show_details(breed):
    breed_name = breed.split("More about ")[-1]
    description = get_dog_description(breed_name)
    return format_description(description, breed_name)

with gr.Blocks(css="""
    .container {
        max-width: 900px;
        margin: 0 auto;
        padding: 20px;
        background-color: rgba(255, 255, 255, 0.9);
        border-radius: 15px;
        box-shadow: 0 0 20px rgba(0, 0, 0, 0.1);
    }
    .gr-form { display: flex; flex-direction: column; align-items: center; }
    .gr-box { width: 100%; max-width: 500px; }
    .output-markdown, .output-image {
        margin-top: 20px;
        padding: 15px;
        background-color: #f5f5f5;
        border-radius: 10px;
    }
    .examples {
        display: flex;
        justify-content: center;
        flex-wrap: wrap;
        gap: 10px;
        margin-top: 20px;
    }
    .examples img {
        width: 100px;
        height: 100px;
        object-fit: cover;
    }
    """) as iface:

    gr.HTML("<h1 style='font-family:Roboto; font-weight:bold; color:#2C3E50; text-align:center;'>🐶 Dog Breed Classifier 🔍</h1>")
    gr.HTML("<p style='font-family:Open Sans; color:#34495E; text-align:center;'>Upload a picture of a dog, and the model will predict its breed, provide detailed information, and include an extra information link!</p>")
    
    with gr.Row():
        input_image = gr.Image(label="Upload a dog image", type="numpy")
        output = gr.Markdown(label="Prediction Results")
    
    with gr.Row():
        btn1 = gr.Button("View More 1", visible=False)
        btn2 = gr.Button("View More 2", visible=False)
        btn3 = gr.Button("View More 3", visible=False)

    input_image.change(predict, inputs=input_image, outputs=[output, btn1, btn2, btn3])
    
    btn1.click(show_details, inputs=btn1, outputs=output)
    btn2.click(show_details, inputs=btn2, outputs=output)
    btn3.click(show_details, inputs=btn3, outputs=output)

    gr.Examples(
        examples=['Border_Collie.jpg', 'Golden_Retriever.jpeg', 'Saint_Bernard.jpeg', 'French_Bulldog.jpeg', 'Samoyed.jpg'],
        inputs=input_image
    )

    gr.HTML('For more details on this project and other work, feel free to visit my GitHub <a href="https://github.com/Eric-Chung-0511/Learning-Record/tree/main/Data%20Science%20Projects/Dog%20Breed%20Classifier">Dog Breed Classifier</a>')

# launch the program
if __name__ == "__main__":
    iface.launch()