jeff's blog

About

Portrait of Jefferson Hernandez

Hola! I am Jefferson Hernandez, a PhD student in Computer Science at Rice University, working in self-supervised learning for multimodal data and reasoning in Large language models under the supervision of Prof. Vicente Ordonez at Vislang Lab.

I am currently a Research Intern at Meta Reality Labs working with Ishwarya Ananthabhotla. I previously interned at Adobe Research working with Kushal Kafle. I have also colaborated with Ruben Villegas on self-supervised learning for video data.

Prior to this, I obtained my bachelor’s degree in Industrial Engineering from ESPOL (top 1% of the class), where I worked with Prof. Andres G. Abad on machine learning and computer vision. I also worked as a research assistant at INARI Lab on applications of computer vision to retail. I have also shortly worked as a Computer Vision engineer at adaviv.

I am interested in computer vision, natural language processing, and machine learning. I am particularly interested in test-time-training for LLMs and self-supervised learning for images, text, video and audio. As a PhD student, I am always eager to collaborate with other researchers. If you are interested in working with me, feel free to reach out to me via email.

πŸ”₯ News


πŸ“ Preprints


GVIT teaser

GViT: Representing Images as Gaussians for Visual Recognition. [arxiv]
Jefferson Hernandez, Ruozhen He, Guha Balakrishnan, Alexander C. Berg, Vicente Ordonez
June 2025

ProxyThinker teaser

ProxyThinker: Test-Time Guidance through Small Visual Reasoners. [arxiv]
Zilin Xiao, Jaywon Koo, Siru Ouyang, Jefferson Hernandez, Yu Meng, Vicente Ordonez
May 2025

GenLLaVA teaser

Generative Visual Instruction Tuning. [arxiv]
Jefferson Hernandez, Ruben Villegas and Vicente Ordonez
June 2024

πŸ“ Publications


cFreD teaser

Evaluating Text-to-Image Synthesis with a Conditional FrΓ©chet Distance [arxiv] [Code]
Jaywon Koo*, Jefferson Hernandez*, Moayed Haji-Ali, Ziyan Yang, Vicente Ordonez
*Equal contribution. March 2025, WACV 2026

cFreD teaser

Improving Large Vision and Language Models by Learning from a Panel of Peers [arxiv]
Jefferson Hernandez, Jing Shi, Simon Jenni, Vicente Ordonez, Kushal Kafle
November 2024, ICCV 2025

ViC-MAE teaser

ViC-MAE: Self-Supervised Representation Learning from Images and Video
with Contrastive Masked Autoencoders. [Paper] [arxiv] [Project Page] [Code]
Jefferson Hernandez, Ruben Villegas and Vicente Ordonez
November 2023, ECCV 2024

Retail Dataset

Automatic Retail Dataset Creation with Multiple Sources of Information Synchronization. [Paper]
Ricardo Palacios, Byron Piguave, Jefferson Hernandez,and Andres Abad
October 2023, IPTA 2023

Action Recognition

A View Invariant Human Action Recognition System for Noisy Inputs. [Paper]
Jefferson Hernandez, J.W. Kim, Ruben Cobos, and Andres Abad
May 2022, CRV 2022

Hierarchical HAR

Hierarchical Human Action Recognition to Measure the Performance of Manual Labor. [Paper]
Jefferson Hernandez, Gabriela Valarezo, Ruben Cobos, J.W. Kim, and Andres Abad
2021, IEEE Access

Time Motion Study

Automatic Time and Motion Study Using Deep Learning. [Paper]
Jefferson Hernandez, Sofia Lopez, Gabriela Valarezo, and Andres Abad
2021, CRC Press

Multi Object Tracking

A fast multi-object tracking system using an object detector ensemble. [Paper] [arxiv]
Jefferson Hernandez, Ruben Cobos, and Andres Abad
June 2019, ColCACI 2019

Retail Traffic Flow

Retail Traffic-Flow Analysis Using a Fast Multi-object Detection and Tracking System. [Paper]
Jefferson Hernandez, Ruben Cobos, and Andres Abad
2019, Springer, ColCACI

RBM Model

Learning from multivariate discrete sequential data
using a restricted Boltzmann machine model. [Paper] [arxiv]
Jefferson Hernandez, and Andres Abad
May 2018, ColCACI 2018

RBM Features

Spatial and Temporal Feature Extraction Using a Restricted Boltzmann Machine Model. [Paper]
Jefferson Hernandez, and Andres Abad
May 2018, Springer, ColCACI