Fabian Perez

AI researcher, i like to train deep neural nets πŸ§ πŸ€–

Bucaramanga, Colombia, GMT-5

FP

About

I am a computer science student at Universidad Industrial de Santander (UIS) in Colombia. I am currently a master student in computer science. I have strong skills in software development and deep learning. My expertise across both these areas allows me to create innovative solutions by bringing them together

Education

Universidad Industrial de Santander

2025 - 2027?
MSc(s) Computer Science

University of Delaware

June 2025 - August 2025
Summer Research Program

King Abdullah University of Science and Technology

September 2024 - February 2025
Research Internship

Universidad Industrial de Santander

2020 - 2024
BSc Computer Science

Interests

Deep Learning
Privacy Preserving Deep Learning
Vision Transformers
Generative AI
Applied Machine Learning
Software Development

Recent Highlights

Papers

UnMix-NeRF: Spectral Unmixing Meets Neural Radiance Fields

Fabian Perez, Sara Rojas Martinez, Carlos Hinojosa, Hoover Rueda-ChacΓ³n, Bernard Ghanem

UnMix-NeRF: Spectral Unmixing Meets Neural Radiance Fields
ICCV 2025

UnMix-NeRF integrates spectral unmixing with NeRFs to enable simultaneous hyperspectral view synthesis and unsupervised material segmentation. Unlike traditional NeRFs that rely solely on RGB, our approach captures the intrinsic spectral properties of materials, allowing precise material segmentation and scene editing.

Neural Radiance Fields
Spectral Unmixing
Hyperspectral Imaging
Material Segmentation

Beyond Appearances: Material Segmentation with Embedded Spectral Information from RGB-D imagery

Fabian perez, Hoover Rueda-Chacon

Beyond Appearances: Material Segmentation with Embedded Spectral Information from RGB-D imagery
CVPR 2024 LatinX Workshop

Pioneered a cutting-edge deep learning framework enhancing material segmentation by embedding spectral data into RGB-D images

Material Segmentation
RGB-D imagery
Multimodal learning

Privacy-Preserving Deep Learning Using Deformable Operators for Secure Task Learning

Fabian perez, Jhon Lopez, Henry Arguello

Privacy-Preserving Deep Learning Using Deformable Operators for Secure Task Learning
ICASSP 2024

we propose a novel Privacy-Preserving framework that uses a set of deformable operators for secure task learning. Our method involves shuffling pixels during the analog-to-digital conversion process to generate visually protected data

Image Privacy
Deformable Operators
Computational Imaging

Semillero Hands-On Computer Vision

Project image

At Hands-on Computer Vision, we're a passionate group of students and researchers from UIS, dedicated to exploring the cutting-edge of computer vision technology. Our journey is a blend of theory and practice

Awards

Google Cloud Vertex AI Agent Builder Hackathon 2024

Yipao

πŸ†First Place
Gallery image 1
Gallery image 2

Revolutionizing business intelligence with AI-driven SQL interactions. Simplify complex queries and enhance database efficiency effortlessly

Indra Hackdar 2024

Hunter

πŸ†First Place
Gallery image 1
Gallery image 2

Artificial Intelligence model-as-a-service tool that facilitates the process of facilitates the process of recruitment and human talent management in companies, and companies,

AMB Geo2Code 2023

DeepBeauty

πŸ†Second Place
Gallery image 1
Gallery image 2
Gallery image 3

My favorite hackathon, a deep learning model for automatic semantic segmentation for aerial imagery in Bucaramanga

Financiera Comultrasan Fedesoft Hackathon 2023

FcPay

πŸ†Second Place
Gallery image 1
Gallery image 2
Gallery image 3

payment code creation and money management application