Welcome, I am

Daniel Cruz

Daniel specializes in

Embedded Systems. AI. Data Science. Remote Sensing. Software Engineering.

Get started

ABOUT ME

AI ENGINEER, DATA SCIENTIST, EMBEDDED SYSTEMS PROGRAMMER

Daniel Cruz is a highly skilled electrical and computer science engineer with a strong academic background, holding a Master of Science in Computer Science and a Bachelor's degree in Electrical Engineering from The University of Texas at El Paso (UTEP), both achieved with a 4.0 GPA. He has extensive experience in embedded systems, microprocessor programming, and data acquisition, having collaborated on various research and development projects.

His skill set includes proficiency in programming languages such as C, C++, and Python, with experience using Linux. He has several years of experience using ML modeling tools such as Tensoflow, PyTorch, and Keras. Daniel is also adept in data science tools like Numpy, Pandas, OpenCV, and Matplotlib. Daniel has also worked with microcontrollers such as Arduino and MSP432, as well as designing PCB layouts using Eagle software.

He has collaborated with several research institutes such as Oak Ridge National Laboratories, Woods Hole Oceanographic Institution (WHOI) and National Oceanic and Atmospheric Administration (NOAA). Daniel's work is international and cross-cultural, having collaborated with companies in Australia and Japan.

Resume

WHAT I DO

Machine Learning

Full ML pipeline from data processing to model deployment.

Data Science

Discover actionable insights. Processing data & building ETL pipelines.

Embedded Systems Design

PCB layout to embedded programming, both hardware and software.

Project showcases

Explore a selection of projects showcasing my skills in AI, data science, and embedded systems.

Thermal Wildlife Detection

Using Computer Vision to Detect and Locate Wildlife

Oil Particle Analysis

Using AI and holography for efficient oil spill analysis.

Whale Detection Project

AI-powered whale detection using aerial imagery.

Image Alignment Project

Analyzing image correspondences and alignment using advanced techniques.

ChatGPT Discord Bot

Integrating ChatGPT into a Discord bot.

Gallery Description

Oil underwater is often in the form of micro-sized oil droplets, making it very difficult to quantify underwater oil spills. Autonomous Underwater Vehicles are being used to monitor oil spills in the ocean. These vehicles use a special camera called the holographic camera to capture images of these oil droplets. Special software can process these images to provide different focal planes from a single image. While useful to capture several several oil droplets at once, it is difficult to measure and quantify the oil droplets because of the processing time and difficulty in processing these images.

My thesis focused on methods to speed up processing these images. Three methods to process these images were developed:

Traditional computer vision technique: This technique used edge detection to find the oil droplets
Semantic Segmentation ML Model: This technique used a U-Net to segment the images based on supervised learning
YOLOv5 Object Detection ML Model: This technique used a YOLO

model to detect the oil droplets and draw a bounding box around the droplet.

Compared to the traditional computer vision technique, the machine learning techniques found more particles and had a speed up of 250% in execution time.

Download Project File

Example of holographic image.

Holographic image reconstructed. Each droplet can be accurately measured on the correct frame.

Segmented image based on tradional computer vision techniques from literature

Segmented image based on a U-net network.

Object Detection based on YOLOv5.