Flask Deep Learning Model Deployment

Project Overview

This Flask project demonstrates the deployment of two deep learning models for different tasks, namely image classification and text classification. The goal is to enable server-side inference through HTTP requests.

Team Members

• Motassim Ahmed Taha

Tasks and Model Selection

Task 1: Image Classification

• Selected Model: ResNet50
• Motivation: ResNet50 is a powerful convolutional neural network (CNN) architecture widely used for image classification. It has shown excellent performance on various image datasets.

Task 2: Text Classification

• Selected Model: BERT (Bidirectional Encoder Representations from Transformers)
• Motivation: BERT is a state-of-the-art transformer-based model known for its superior performance in natural language processing tasks. It excels in capturing contextual information and relationships within text.

Model Integration

We ensured correct usage of the selected models by following these steps:

1. Model Selection: We chose models that are pre-trained on large datasets and suitable for the respective tasks.
2. Testing: We performed testing on sample data to ensure the models provide accurate predictions.
3. Compatibility: Ensured that the models are compatible with the libraries and frameworks used for deployment (e.g., PyTorch, Hugging Face).

Deployment

The models are deployed using Flask, PyTorch, and Hugging Face. The server enables inference through POST HTTP requests and returns results in JSON format.

Endpoints

• Image Classification Endpoint: /predict_image
• Text Classification Endpoint: /predict_text

Example Usage

Image Classification

curl -X POST -H "Content-Type: application/json" -d '{"image_url": "https://example.com/image.jpg"}' http://localhost:5000/predict_image

Web Interface

We have developed a user-friendly web interface to illustrate the use of the server. The interface includes the following routes:

• /: Home page with project overview and instructions.

• /image_classification: Interface for image classification with an option to upload an image or provide a URL.

• /text_classification: Interface for text classification with a text input box.

Feel free to explore the web interface and experience the seamless integration of deep learning models for image and text classification.

Getting Started

To run the project locally, follow these steps:

1. Clone the repository:

git clone https://github.com/yourusername/flask-deep-learning.git
cd flask-deep-learning

2. Install dependencies:

pip install -r requirements.txt

4. Access the web interface at http://localhost:5000 and start testing the models.