Streamlined DETR Object Detection Model
A reimplementation of the DETR model, focusing on architectural optimizations and learned positional encodings to enhance training efficiency and inference speed.
A reimplementation of the DETR model, focusing on architectural optimizations and learned positional encodings to enhance training efficiency and inference speed.
A custom npm-style registry tailored for ACME Corporation's needs, featuring secure package ingestion, dependency analysis, and version management. Enhanced with RESTful APIs and a user-friendly interface for efficient package handling.
Analyzed bicycle traffic data from NYC bridges using Python to optimize sensor placement, explore weather-based rider prediction, and assess traffic patterns by day. Developed models and visualizations to enhance urban planning and bike safety.
Algorithm utilizing OpenCV and Matplotlib in Python to process MP4 video files, extracting the dominant color from each frame and generating a visual representation. This tool aids in color pattern analysis and visualization for video content.