Anirudh Mani
Built and trained a custom neural network to extract handwritten characters from images with over 92% test accuracy
Built a real-time AR system that detects and matches visual features to overlay graphics on live video using homography estimation and image warping
Implemented an Encoder-Decoder model and fine-tuned hyperparameters to optimize classification performance
Interacts with a language model API to identify entities in text, while exploring the impact of different demonstration modes and shot variations on model performance
Designed and developed an app for creating, managing, and competing with fantasy soccer teams
Conducted in-depth analysis of cache line management and quantified the impact of various operations on cache performance