Varadh Kaushik's repositories
RL-models-for-Super-Mario-Bros
This repository explores the application of three reinforcement learning algorithms—Deep Q-Networks (DQN), Double Deep Q-Networks (DDQN), and Proximal Policy Optimization (PPO)—for playing Super Mario Bros using the OpenAI Gym and nes-py emulator. It includes a comparative analysis of these models.
AI-Character-Forge
Advanced AI-driven tool for generating unique video game characters using Stable Diffusion, DreamBooth, and LoRa adaptations. Enhances creativity with customizable, high-quality character designs, tailored specifically for game developers and artists.
Autonomous-Kinect-Robot
An advanced robotics initiative harnessing Kinect's computer vision capabilities for precise object detection and manipulation, coupled with robust autonomous navigation and path planning technologies. Powered by an Arduino for seamless hardware control and real-time responsiveness.
CassavaHealth-AI
Based on my published research paper, this project uses a "One-vs-All" deep learning approach with EfficientNet B4 to classify cassava leaf diseases. Integrated into an Android app, it helps farmers detect diseases early, supporting sustainable farming and reducing crop losses.
NLP-SoundSculptor
This project merges Natural Language Processing with music generation, using NLP techniques to analyze lyrics and guide the creation of music. It employs sentiment analysis, keyword extraction, and topic modeling, along with Genius and Spotify APIs, to generate music that reflects the emotional depth of the lyrics using the MusicGen model.
Hazard-View-Bird
Efficient model for semantic segmentation on edge devices, specifically targeting the analysis of disaster scenes from images captured by unmanned aerial vehicles (UAVs).
trade-recommendations
This repository contains the implementation of a transformer-based model combined with a Proximal Policy Optimization (PPO) model to generate trade recommendations. The project leverages the predictive capabilities of transformers for price forecasting and the strategic decision-making of reinforcement learning.
travel-planner
Leveraging advanced AI to dynamically create personalized travel itineraries, this tool utilizes a Large Language Model, real-time web scraping, and Retrieval-Augmented Generation to adapt travel plans instantly based on user preferences and live data updates.