Emoji Fi-Gen

Introduction

This project provides an interactive interface to find and generate emojis based on input text. The interface consists of an input box for entering text and two buttons: Find and Generate. When a user submits text, the system processes it through a BERT model and calculates cosine similarity to match emojis from a predefined dataset. If matching emojis are found, the top five emojis are displayed. By clicking the Generate button, the user can generate an emoji using an API call to a fine-tuned model on a specific emoji dataset.

Features

• Text Input: Input box to enter text for emoji search.
• Submit Button: Processes the input text and displays matching emojis.
• Generate Button: Generates an emoji using a fine-tuned model via API call.

Installation

Prerequisites

• streamlit
• replicate
• pillow
• transformers
• gensim==3.8.3
• emoji-data==0.1.6
• scikit-learn==1.4.2
• scipy==1.10.0

Setup

1. Clone the Repository

   git clone https://github.com/Lakshyadevelops/EmojiGen.git

2. Create a virtual environment

   conda create -n emojigen python=3.9
   conda activate emojigen

3. Install Dependencies

   pip install -r requirements.txt

4. Run the Streamlit Application

   streamlit run main.py

Usage

1. Open the Application Open your web browser and go to http://localhost:8501.

2. Input Text Enter the text you want to analyze in the input box.

3. Find Text Click the Find button. The text will be processed, and the top four matching emojis will be displayed if found.

4. Generate Emoji Click the Generate button to generate an emoji using the fine-tuned model via API call.

How It Works

1. Text Processing

   • The input text is passed through a pre-trained BERT model to obtain embeddings.

2. Cosine Similarity Calculation

   • Cosine similarity is calculated between the text embeddings and the embeddings of emojis in the dataset.

3. Emoji Matching

   • The top five matching emojis based on cosine similarity are displayed.

4. Emoji Generation

   • Upon clicking the Generate button, an API call is made to a server where a fine-tuned model generates an emoji based on the input text.

Contribution

Feel free to fork this repository and contribute by submitting a pull request.

License

This project is licensed under the MIT License.

---