Computer Vision · Event-based Vision · Data Science

NICOLAS CARRASCO

Advanced Computer Science student at University of Buenos Aires and Research Intern at the Image, Vision and Intelligent Systems Lab (LIVSI) at Instituto de Ciencias de la Computación (CONICET-UBA).

Office 2115, Pabellón 0+inf, Ciudad Universitaria, Buenos Aires, Argentina

Me.

01

Research

Event-based vision for eye movement analysis

Real-time eye tracking with Dynamic Vision Sensors, exploring high-frequency pupil position estimation, blink detection, and gaze estimation from sparse temporal events.

Current work spans hybrid event-plus-frame pipelines and event-only methods, including parametric pupil models and real-time processing constraints in non-specialized hardware.

02

Publications

Aug 2026

High-frequency event-based Eye Tracking

Nicolas Carrasco, Nicolas Mastropasqua, Daniel Acevedo

55th Jornadas Argentinas de Informática (JAIIO) In Press

  • Event-based vision
  • Eye tracking
  • Computer vision

03

Selected Projects

2025

Supervised Learning for Cancer Prognosis Prediction

Compared supervised classifiers for prognosis prediction after exploratory data analysis and class-imbalance analysis. Evaluation used a custom stratified train/test split and stratified five-fold cross-validation.

Decision Tree, KNN, SVM, LDA, and Naive Bayes models were tuned with Random Search and assessed with AUROC, AUPRC, and accuracy. SVM was selected while prioritizing clinically meaningful operating points and reduced false negatives.

Python · scikit-learn · Model selection

04

Experience

Sep 2025 - Present

Research Intern

ICC — UBA / CONICET

  • Research in real-time eye tracking methods using event cameras.
  • Develop and evaluate approaches for pupil tracking, blink detection, and gaze estimation.
  • Work across event-based vision, machine learning, deep learning, and real-time systems.

05

Education

2021 - 2027

Bsc+Msc in Computer Science

University of Buenos Aires (UBA)