CODSOFT AI Internship – Project Repository

This repository contains all 5 AI/ML projects completed as part of the CODSOFT Artificial Intelligence Internship.
Each task demonstrates core AI concepts, including rule-based automation, ML model training, computer vision, NLP, and recommender systems.

---

Project Structure

CODSOFT/
 ├── task1_rule_based_chatbot/
 ├── task2_tictactoe_ai/
 ├── task3_image_captioning/
 ├── task4_movie_recommendation/
 ├── task5_face_detection/
 └── README.md

---

TASK 1 — Rule-Based Chatbot

Objective

Build a simple rule-based chatbot that responds to predefined patterns using conditional logic, keyword detection, or simple NLP preprocessing.

Features

• Pattern-based responses
• Basic NLP (lowercasing, keyword checks, fallback responses)
• Can be extended with more rules

How to Run

python chatbot.py

---

TASK 2 — Tic Tac Toe AI (Minimax Algorithm)

Objective

Build an unbeatable Tic Tac Toe bot using the Minimax algorithm.

Features

• Human vs AI gameplay
• AI always plays optimal moves
• Minimax with recursion & evaluation scores

How to Run

python tictactoe.py

---

TASK 3 — Image Captioning (CNN + LSTM Model)

Objective

Generate captions for images using a pre-trained CNN encoder and an LSTM decoder.

Features

• Uses InceptionV3 / ResNet for feature extraction
• Decodes captions using a trained LSTM
• Works on any input image
• Fully modular: preprocessing → encoding → decoding

How to Run

python app.py

---

TASK 4 — Movie Recommendation System

Objective

Build a movie recommender using NLP-based similarity.

Features

• Content-based similarity
• TF-IDF vectorization or cosine similarity
• The user can search for any movie
• Returns top similar movies
• Optional add-ons completed (sorting, add runtime movies, etc.)

How to Run

python recommend.py

---

TASK 5 — Face Detection + Recognition App

Objective

Develop a real-time face detection and recognition system using Haar cascades and LBPH.

Features

• Face detection using Haar models
• Face recognition using LBPH
• Dataset collection through webcam
• Model training & testing scripts included
• Compatible with lightweight systems

How to Run

1️ Collect Faces

python capture_faces.py

2️ Train LBPH Model

python train.py

3️ Recognize Faces in Real-Time

python recognize.py

---

Technologies Used

• Python
• OpenCV
• Numpy
• Pillow
• TensorFlow / Keras (Image Captioning)
• NLP (TF-IDF, cosine similarity)
• Minimax algorithm
• Rule-based logic

---

Notes

• Each project is completely modular and can run independently.
• Requirements for each task are inside the respective folders.
• No external images/screenshots included as per project requirement.

---

Author

Aneesh Srinivas
CODSOFT Artificial Intelligence Internship – 2025

---

If you like the project, consider giving the repo a star!