Nerv Analytica

by Connor Finnerty

Specializing in neuroscience, data science, optical physics, and biomedical sensing.

Portfolio

A selection of my work in neuroscience, machine vision, and biomedical engineering.

Breathing Rate Estimation

Applied signal processing to extract breathing frequency from cardiac intervals using FFT and spectral refinement.

Technical Details:

Signal ProcessingSpectral EstimationFourier AnalysisSciPyCardiopulmonary Physiology

NBA Player Identifier

YOLOv8 + CNN + OCR pipeline that detects Celtics players in images and identifies them by jersey number using Basketball Reference data.

Technical Details:

PythonYOLOv8TensorFlowEasyOCR

Election Donation Analysis

Linked FEC campaign donations with open voter files to analyze demographic and ideological patterns in political contributions.

Technical Details:

PythonPandasRegexGeopandas

Reddit Personality + Ideology NLP

NLP pipeline that analyzes Reddit comment histories to infer psychological traits and ideological leanings using TF-IDF and clustering.

Technical Details:

PythonPRAWSpaCyScikit-learn

NV-Diamond Magnetometry

Built an optical detection system using nitrogen-vacancy centers in diamond to sense neuron-level magnetic activity. Includes CNN signal modeling.

Technical Details:

Lab InstrumentationCNN ModelingHarvard

Laser Imaging with CNNs

Developed a CNN-based detection system for off-axis laser scatter under DARPA funding. Integrated optics design and neural image segmentation.

Technical Details:

CNNsLaser OpticsDARPAUSC

Technical Toolbox

A curated overview of the tools and domains I work in—spanning data science, machine learning, and experimental physics.

Data Science & Machine Learning

  • Languages: Python, TypeScript, C++, MATLAB
  • Frameworks: PyTorch, TensorFlow, scikit-learn, MONAI, Hugging Face
  • Libraries: Pandas, NumPy, OpenCV, Matplotlib, Seaborn
  • Specialized: YOLOv8, EasyOCR, MNE-Python, Kilosort
  • Techniques: Deep learning, OCR pipelines, signal processing, forecasting
  • Deployment: Streamlit, Flask, FastAPI, Docker

Experimental Science & Instrumentation

  • Domains: NV-magnetometry, EEG, eye-tracking, biomedical sensing
  • Institutions: Harvard, USC (DARPA), Smithsonian
  • Hardware: Arduino, DAQ, Thorlabs optics, 3D printing
  • Tools: MATLAB, LabView, CMOS/CCD imaging arrays
  • Techniques: Spike sorting, signal amplification, multimodal analysis

Get in Touch

Whether for collaborations, inquiries, or simply to connect, I'm always open to discussion.