Update README.md

README.md

@@ -1,20 +1,146 @@
+# 🧠 SmartVision AI  
+### Intelligent Multi-Class Object Recognition System
+
+SmartVision AI is an **end-to-end Computer Vision application** that performs  
+**image classification**, **object detection**, and **real-time inference** using
+state-of-the-art deep learning models.  
+The project demonstrates the complete AI lifecycle — from **model training** to
+**optimized deployment** using **Streamlit**.
+
+---
+
+## 🚀 Key Features
+
+- 🖼️ **Image Classification**
+  - Custom-trained deep learning models
+  - Top-5 prediction display with confidence scores
+  - Side-by-side comparison of multiple CNN architectures
+
+- 📦 **Object Detection**
+  - Pretrained YOLOv8 model for real-time object detection
+  - Bounding boxes, class labels, and confidence scores
+  - Adjustable confidence threshold
+
+- 📸 **Live Webcam Detection (Optimized)**
+  - Real-time detection using webcam
+  - FPS monitoring and CPU-friendly optimizations
+  - Frame skipping and resolution scaling
+
+- 📊 **Model Performance Dashboard**
+  - Accuracy comparison (Train / Validation / Test)
+  - Inference speed analysis
+  - Visual performance insights
+
+- ⚡ **Optimized Inference**
+  - Lightweight models for CPU execution
+  - Streamlit caching for faster loading
+  - Performance-focused design decisions
+
+---
+
+## 🏗️ Model Architectures Used
+
+### 🔹 Image Classification
+- **VGG16 (Custom Trained)**
+- **ResNet50**
+- **MobileNetV2**
+- **EfficientNet-B0**
+
+### 🔹 Object Detection
+- **YOLOv8 (Pretrained on COCO Dataset)**
+
+---
+
+## 📂 Dataset Information
+
+- **Image Classification Dataset**
+  - Domain-specific dataset
+  - 25 object classes
+  - Train / Validation / Test split
+  - Image preprocessing and augmentation applied
+
+- **Object Detection Dataset**
+  - COCO Dataset
+  - 80 general-purpose object classes
+  - Bounding box annotations
+
+---
+
+## 🛠️ Tech Stack
+
+**Programming Language**
+- Python 🐍
+
+**Deep Learning & Computer Vision**
+- PyTorch
+- Torchvision
+- Ultralytics YOLOv8
+- OpenCV
+
+**Data Analysis & Visualization**
+- NumPy
+- Pandas
+- Matplotlib
+- Seaborn
+
+**Web & Deployment**
+- Streamlit
+- VS Code
+- Git & GitHub
+
 ---
-title: Smartvision Ai
-emoji: 🚀
-colorFrom: red
-colorTo: red
-sdk: docker
-app_port: 8501
-tags:
-- streamlit
-pinned: false
-short_description: end to end computer vision system
-license: mit
+
+## ⚡ Performance Optimization Techniques
+
+- Frame skipping for real-time webcam inference
+- Reduced image resolution for faster detection
+- Lightweight YOLOv8n model for CPU execution
+- Streamlit resource caching
+- Confidence-based filtering of predictions
+
 ---
 
-# Welcome to Streamlit!
+## 📁 Project Structure
+SmartVisionAI/
+│
+├── app.py # Main Streamlit application
+├── requirements.txt # Python dependencies
+├── README.md # Project documentation
+├── Image.txt/ # Images, icons, logos
+├── yolo.ipynb
+└── smartvisionAI.ipynb(Downloading and training process of data)
+
+## Screenshots
+<img width="1920" height="1080" alt="Screenshot (176)" src="https://github.com/user-attachments/assets/feb97730-b862-4504-ac37-bc733fe21aba" />
+
+<img width="1920" height="1080" alt="Screenshot (178)" src="https://github.com/user-attachments/assets/60abc70b-3d2c-4aee-aaf2-53dade77d7e3" />
+Demo Images
+
+<img width="1920" height="1080" alt="Screenshot (177)" src="https://github.com/user-attachments/assets/5c03e4b4-eaa5-4eb8-942f-b3fae16db210" />
+Detection
+
+<img width="1253" height="825" alt="Screenshot 2025-12-14 at 08-17-06 SmartVision AI - Intelligent Multi-Class Object Recognition System" src="https://github.com/user-attachments/assets/158bdfca-c160-4968-a508-d3cd47878768" />
+
+
+
+
+## 📌 Note on Model Files
+
+-Due to size constraints, trained model weights (.pt, .pth) are not included
+in this repository.
+
+## 🎓 Academic & Practical Relevance
+
+- This project was built to:
+
+- Demonstrate practical Deep Learning & Computer Vision skills
+
+- Showcase model deployment and optimization
+
+- Serve as a portfolio project for interviews and evaluations
 
-Edit `/src/streamlit_app.py` to customize this app to your heart's desire. :heart:
+## 👨‍💻 Developer
 
-If you have any questions, checkout our [documentation](https://docs.streamlit.io) and [community
-forums](https://discuss.streamlit.io).
+Rahul Kumar
+B.Tech in Information Technology
+IIEST Shibpur