done

.gitignore

@@ -0,0 +1,2 @@
+.env
+cnn_model1.pth

.streamlit/config.toml

@@ -0,0 +1,7 @@
+[theme]
+base="light"
+primaryColor="#0d6efd"
+backgroundColor="#0f1214"
+secondaryBackgroundColor="#FFFFFF"
+textColor="#31333F"
+font="sans serif"

README.md

@@ -1,6 +1,6 @@
 ---
 title: Tumor Prediction
-emoji: 💻
+emoji: 🧠
 colorFrom: pink
 colorTo: blue
 sdk: streamlit
@@ -9,4 +9,29 @@ app_file: app.py
 pinned: false
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+
+# Tumor Prediction Application
+
+![Tumor Prediction Preview](preview.png)
+
+## Overview
+
+This application uses advanced machine learning models to predict the type of tumor from medical images. It leverages the latest advancements in artificial intelligence to provide accurate and quick predictions.
+
+## Features
+
+- **Image Upload**: Upload an image of a tumor for prediction.
+- **Prediction**: Get a prediction of the tumor type with confidence percentage.
+- **Translation**: Translate the prediction summary from French to Spanish.
+- **Text-to-Speech**: Convert the translated text to speech.
+- **Interactive UI**: User-friendly interface with Bootstrap and Particles.js integration.
+
+## Technologies Used
+
+- **Streamlit**: For building the web application.
+- **PyTorch**: For the deep learning model.
+- **Hugging Face API**: For text summarization and translation.
+- **Google Text-to-Speech (gTTS)**: For converting text to speech.
+- **Mistral LLM**: For generating detailed medical diagnoses.
+- **Bootstrap**: For responsive design.
+- **Particles.js**: For background particle effects.

app.py

@@ -0,0 +1,74 @@
+from functions import generate_text, get_diagnosis, summarize_text, text_to_speech, translate_fr_to_es
+from predict import predict
+from utils import load_image
+from dotenv import load_dotenv
+import streamlit as st
+from layout import show_header, show_prediction_section, show_footer
+from styles import inject_particles_js, inject_bootstrap_and_custom_css
+
+# Configuration de la page
+st.set_page_config(
+    page_title="Application Multitâche",
+    page_icon=":robot_face:",
+    layout="wide",
+    initial_sidebar_state="auto",
+)
+
+load_dotenv()
+
+# Injecter les styles et scripts
+inject_particles_js()
+inject_bootstrap_and_custom_css()
+
+# Charger les images
+image_path = "images/tumor.png"
+image_base64 = load_image(image_path)
+linkedin_image_base64 = load_image("images/skill-icons--linkedin.svg")
+github_image_base64 = load_image("images/skill-icons--github-dark.svg")
+huggingface_image_base64 = load_image("images/Hugging_Face_idJ6-I79C__1.svg")
+favicon_image_base64 = load_image("images/favicon.png")
+
+# Afficher l'en-tête
+show_header(favicon_image_base64, image_base64)
+
+# Afficher la section de prédiction
+show_prediction_section()
+
+_, col2, col3, _ = st.columns([2, 6, 6, 2])
+
+translation = None
+with col2:
+    uploaded_file = st.file_uploader("Téléchargez une image de tumeur...", type=["jpg", "jpeg", "png"])
+    if uploaded_file:
+        _, col2, _ = st.columns([1, 3, 1])
+        with col2:
+            st.image(uploaded_file, caption="Image de tumeur", width=400)
+            button = st.button("Prédire", use_container_width=True)
+            if button:
+                st.spinner("Prédiction en cours...")
+                predicted, confidence = predict(uploaded_file)
+                diagnosis_prompt = get_diagnosis(predicted, confidence)
+                diagnosis = generate_text(diagnosis_prompt)
+                summary = summarize_text(diagnosis)
+                translation = translate_fr_to_es(summary)
+                output_path = text_to_speech(translation)
+
+if uploaded_file and translation:
+    with col3:
+        st.markdown(
+            f"""
+            <div class="card border-0 shadow-sm p-4">
+                <div class="card-body">
+                    <h2 class="gradient-text"> Resultat de la prédiction </h2>
+                    <p class="lead"> {summary} </p>
+                    <p class="lead"> {translation} </p
+                </div>
+            </div>
+            """,
+            unsafe_allow_html=True, 
+        )
+
+        st.audio(output_path, format="audio/mp3")
+
+# Afficher le pied de page
+show_footer(linkedin_image_base64, github_image_base64, huggingface_image_base64)

functions.py

@@ -0,0 +1,72 @@
+import requests
+import os
+import time
+from gtts import gTTS
+
+
+HF_API_KEY = os.getenv("HUGGING_FACE_API_KEY")
+MISTRAL_API_KEY = os.getenv("MISTRAL_IA_API_KEY")
+
+
+# Summarisation model
+def query(payload):
+    headers = {"Authorization": f"Bearer {HF_API_KEY}"}
+    API_URL_SUMMARISATION_MODEL = "https://api-inference.huggingface.co/models/facebook/bart-large-cnn"
+    response = requests.post(API_URL_SUMMARISATION_MODEL, headers=headers, json=payload)
+    return response.json()[0]['summary_text']
+
+def summarize_text(text):
+    payload = {"inputs": text}
+    return query(payload)
+
+#Text generation model
+def get_diagnosis(prediction, confidence):
+    labels = ["glioma", "meningioma", "notumor", "pituitary"]
+    result = labels[prediction]
+    prompt = f'Tu es un médecin qui doit annoncer à un patient le type de tumeur qu\'il a. Le type de tumeur prédit est "{result}" avec une confiance de {confidence}. Veuillez fournir un diagnostic concis, précis et des conseils.'
+    return prompt
+
+
+def generate_text(input_text, max_retries=3):
+    MISTRAL_API_URL = "https://api.mistral.ai/v1/chat/completions"
+    
+    headers = {
+        "Authorization": f"Bearer {MISTRAL_API_KEY}",
+        "Content-Type": "application/json"
+    }
+    
+    payload = {
+        "model": "mistral-small-latest",
+        "messages": [{"role": "user", "content": input_text}]
+    }
+
+    for _ in range(max_retries):
+        response = requests.post(MISTRAL_API_URL, headers=headers, json=payload)
+        if response.status_code == 200:
+            return response.json()['choices'][0]['message']['content']
+        else:
+            time.sleep(5)
+
+    return "Failed to generate text after multiple attempts"
+
+# Translation model
+def translate_fr_to_es(input_text, max_retries=5):
+    API_URL = "https://api-inference.huggingface.co/models/Helsinki-NLP/opus-mt-fr-es"
+    headers = {"Authorization": f"Bearer {HF_API_KEY}"}
+    payload = {"inputs": input_text}
+
+    for _ in range(max_retries):
+        response = requests.post(API_URL, headers=headers, json=payload)
+        if response.status_code == 200:
+            return response.json()[0]['translation_text']
+        else:
+            time.sleep(5)
+
+    return "Une erreur s'est produite veuillez reessayer plus tard."
+
+# Audio to text model
+def text_to_speech(input_text):
+    outputh_path = "text_to_speech.mp3"
+    tts = gTTS(text=input_text, lang="es")
+    tts.save(outputh_path)
+    return outputh_path

layout.py

@@ -0,0 +1,69 @@
+import streamlit as st
+
+def show_header(favicon_image_base64, image_base64):
+    st.markdown(
+        f"""
+        <div class="container card border-0 shadow-sm p-4 ">
+            <div class="card-body">
+                <div class="row d-flex">
+                    <div class="col-md-6 position-relative">
+                        <img src="data:image/png;base64,{favicon_image_base64}" alt="favicon" class="position-absolute w-100 h-auto text-center top-0 bottom-0 z-0">
+                        <h1 class="gradient-text"> Prédiction de la Tumeur avec l'Intelligence Artificielle </h1>
+                        <h4 style="text-align: justify"> Utilisez notre solution avancée pour prédire le type de tumeur à partir d'images médicales. Téléchargez une image de tumeur et obtenez une prédiction précise en quelques secondes. </h4>
+                    </div>
+                    <div class="col-md-6 d-flex border-0">
+                        <img src="data:image/png;base64,{image_base64}" class="img-fluid mx-auto" alt="Responsive image">
+                    </div>
+                </div>
+            </div>
+        </div>
+        """,
+        unsafe_allow_html=True,
+    )
+
+def show_prediction_section():
+    st.markdown(
+        f"""
+        <div class="container mb-64" id="competences">
+            <div class="h1 text-center my-4 gradient-text">
+                Faites une prédiction
+            </div>
+        </div>
+        """,
+        unsafe_allow_html=True,
+    )
+
+def show_footer(linkedin_image_base64, github_image_base64, huggingface_image_base64):
+    st.markdown(
+        f"""
+         <footer class="container shadow-sm text-center text-lg-start mt-5">
+                <div class="container p-4">
+                    <div class="row">
+                        <div class="col-lg-6 col-md-12 mb-4 mb-md-0">
+                            <h5 class="text-uppercase gradient-text">Contact Information</h5>
+                            <p>
+                                Yannick Simo<br>
+                                Junior Data Scientist<br>
+                                Email: [email protected]<br>
+                            </p>
+                        </div>
+                        <div class="col-lg-6 col-md-12 mb-4 mb-md-0">
+                            <h5 class="text-uppercase gradient-text">Suivez-moi</h5>
+                            <a href="https://www.linkedin.com/in/simo-yannick-38137b231/" class="btn btn-floating p-0" role="button">
+                                <img src="data:image/svg+xml;base64,{linkedin_image_base64}" class="w-50" alt="linkedin icon">
+                            </a>
+                            <a href="https://github.com/Skym616" class="btn btn-floating p-0" role="button">
+                                <img src="data:image/svg+xml;base64,{github_image_base64}" class="w-50" alt="github icon">
+                            </a>
+                            <a href="https://huggingface.co/Skym616" class="btn btn-floating p-0" role="button">
+                                <img src="data:image/svg+xml;base64,{huggingface_image_base64}" class="w-50" alt="huggingface icon">
+                            </a>
+                    </div>
+                    </div>
+                </div>
+                <div class="text-center p-3 text-white gradient-text">
+                    &copy; 2025 Yannick Simo. All rights reserved.
+                </div>
+            </footer>
+        """, unsafe_allow_html=True
+    )

predict.py

@@ -0,0 +1,73 @@
+import torch
+import torch.nn as nn
+import torchvision.transforms as transforms
+from PIL import Image
+
+# Architecture du modèle
+class DeepCNN(nn.Module):
+    def __init__(self, num_classes=4):
+        super(DeepCNN, self).__init__()
+        self.layer1 = nn.Sequential(
+            nn.Conv2d(3, 32, kernel_size=3, padding=1),
+            nn.BatchNorm2d(32),
+            nn.ReLU(),
+            nn.MaxPool2d(kernel_size=2)
+        )
+        self.layer2 = nn.Sequential(
+            nn.Conv2d(32, 64, kernel_size=3, padding=1),
+            nn.BatchNorm2d(64),
+            nn.ReLU(),
+            nn.MaxPool2d(2)
+        )
+        self.layer3 = nn.Sequential(
+            nn.Conv2d(64, 128, kernel_size=3),
+            nn.BatchNorm2d(128),
+            nn.ReLU(),
+            nn.MaxPool2d(2)
+        )
+        self.lqyer4 = nn.Sequential(
+            nn.Conv2d(128, 256, kernel_size=3),
+            nn.BatchNorm2d(256),
+            nn.ReLU(),
+            nn.MaxPool2d(2)
+        )
+        self.fc_layers = nn.Sequential(
+            nn.Linear(28800, 1024),
+            nn.ReLU(),
+            nn.Linear(1024, num_classes)
+        )
+
+    def forward(self, x):
+        out = self.layer1(x)
+        out = self.layer2(out)
+        out = self.layer3(out)
+        out = out.view(out.size(0), -1)
+        out = self.fc_layers(out)
+        return out
+
+def load_model(model_path, num_classes=4):
+    model = DeepCNN(num_classes=num_classes)
+    model.load_state_dict(torch.load(model_path, map_location=torch.device('cpu')))
+    model.eval()
+    return model
+
+# Charger le modèle
+model = load_model('cnn_model1.pth')
+
+# Définir les transformations
+transform = transforms.Compose([
+    transforms.Resize((128, 128)),
+    transforms.ToTensor(),
+    transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])
+])
+
+# Fonction de prédiction
+def predict(image):
+    image = Image.open(image)
+    image = transform(image).unsqueeze(0)
+    with torch.no_grad():
+        outputs = model(image)
+        probabilities = torch.nn.functional.softmax(outputs, dim=1)
+        confidence, predicted = torch.max(probabilities, 1)
+        print(predicted.item(), confidence.item() * 100)
+    return predicted.item(), round(confidence.item() * 100, 2)

requirements.txt

@@ -0,0 +1,7 @@
+gTTS==2.5.4
+Pillow==11.1.0
+python-dotenv==1.0.1
+Requests==2.32.3
+streamlit==1.42.0
+torch==2.6.0+cu126
+torchvision==0.21.0+cu126

styles.py

@@ -0,0 +1,79 @@
+import streamlit as st
+import streamlit.components.v1 as components
+
+def inject_particles_js():
+    particles_html = """
+        <div id="particles-js" style="position: fixed; width: 100%; height: 100%; z-index: -1;"></div>
+        <script src="https://cdn.jsdelivr.net/particles.js/2.0.0/particles.min.js"></script>
+        <script>
+            particlesJS("particles-js", {
+                "particles": {
+                    "number": { "value": 100, "density": { "enable": true, "value_area": 800 } },
+                    "color": { "value": "#f3ec78" },
+                    "shape": { "type": "circle" },
+                    "opacity": { "value": 0.5, "random": false },
+                    "size": { "value": 3, "random": true },
+                    "line_linked": { "enable": true, "distance": 150, "color": "#f3ec78", "opacity": 0.4, "width": 1 },
+                    "move": { "enable": true, "speed": 2, "direction": "none", "random": false, "straight": false, "out_mode": "out" }
+                },
+                "interactivity": {
+                    "detect_on": "canvas",
+                    "events": { "onhover": { "enable": true, "mode": "repulse" }, "onclick": { "enable": true, "mode": "push" } },
+                    "modes": { "repulse": { "distance": 100 }, "push": { "particles_nb": 4 } }
+                }
+            });
+        </script>
+    """
+    components.html(particles_html, height=1900, scrolling=True)
+
+def inject_bootstrap_and_custom_css():
+    st.markdown(
+        """
+        <link href="https://cdn.jsdelivr.net/npm/[email protected]/dist/css/bootstrap.min.css" rel="stylesheet" integrity="sha384-QWTKZyjpPEjISv5WaRU9OFeRpok6YctnYmDr5pNlyT2bRjXh0JMhjY6hW+ALEwIH" crossorigin="anonymous">
+        <script src="https://cdn.jsdelivr.net/npm/[email protected]/dist/js/bootstrap.bundle.min.js" integrity="sha384-YvpcrYf0tY3lHB60NNkmXc5s9fDVZLESaAA55NDzOxhy9GkcIdslK1eN7N6jIeHz" crossorigin="anonymous"></script>
+        <link href="https://fonts.googleapis.com/css2?family=Poppins:wght@300;400;600;700&display=swap" rel="stylesheet">
+        <style>
+            body, html, * {
+                font-family: 'Poppins', sans-serif;
+            }
+            body {
+                background-color: #0c1926;
+            }
+            .card, footer{
+                background-color: #0c1926;
+                color: #ffffff;
+            }
+            .stIFrame{
+                position: absolute;
+            }
+            h1, h2 {
+                font-size: 2.5rem;
+            }
+            p, span, div, a, button {
+                font-weight: 400;
+            }
+            .stMainBlockContainer{
+                padding-top: 48px !important;
+            }
+            p , .stFileUploaderFileName, small{
+                font-weight: 600;
+                color: #ffffff !important;
+            }
+            section[data-testid="stFileUploaderDropzone"]{
+                color: #ffffff !important;
+                background-color: #0c1926 !important;
+            }
+            .gradient-text {
+                background-color: red;
+                background-image: linear-gradient(45deg, #f3ec78, #af4261);
+                background-size: 100%;
+                background-repeat: repeat;
+                -webkit-background-clip: text;
+                -webkit-text-fill-color: transparent; 
+                -moz-background-clip: text;
+                -moz-text-fill-color: transparent;
+            }
+        </style>
+        """,
+        unsafe_allow_html=True,
+    )

utils.py

@@ -0,0 +1,12 @@
+import base64
+import requests
+
+def load_image(image_path):
+    with open(image_path, "rb") as img_file:
+        return base64.b64encode(img_file.read()).decode()
+
+def load_lottie_url(url):
+    r = requests.get(url)
+    if r.status_code != 200:
+        return None
+    return r.json()