This project combines sentiment analysis with image captioning using the Salesforce BLIP model and Hugging Face's Transformers library. The chatbot analyzes text input sentiment and provides captions for uploaded images.
[INPUT PROCESSING WINDOW]
Python: Backend development and machine learning. Flask: Web framework for handling backend requests. Streamlit: Tool for deploying and serving the web application. Hugging Face Transformers: Library for NLP and image captioning. HTML, CSS, JavaScript: Frontend development for user interaction.
Install Dependencies: pip install transformers torch torchvision flask streamlit pillow
Backend:
python app.py Frontend (Streamlit):
streamlit run frontend.py Accessing the Application:
Open your browser and navigate to http://localhost:8501 to interact with the chatbot and image captioning interface.
app.py: Flask backend for processing requests. frontend.py: Streamlit frontend for user interface. requirements.txt: List of Python dependencies.
Sentiment Analysis:
Enter text in the chat interface and receive sentiment analysis results.
Upload an image to get automated captions generated by the BLIP model.
[RESULT GENERATION WINDOW]
[EXPLAINS VARIUOUS FUNCTIONBALITY OF ADDING AN IMAGE THROUGH WEBCAM, SYSTEM IMAGES AND SO ON]
Deployed using GRADIO for easy web deployment and sharing (Streamlit/Flask or other cloud platforms can also be used). Please find the deployed link of web application in description for testing/implementation purposes