Fredy Barrientos-Espillco

Research Scientist & AI Engineer | Deep Learning · Computer Vision · Generative Models

Email: fredybar[at]ucm[dot]es
Phone: (+34) 91 394 4375
Room: Physics - 237

Bio

Hi! I’m Fredy Barrientos-Espillco, Ph.D. in Computer Science and Research Scientist in Artificial Intelligence.

I am a research scientist and engineer specialized in Deep Learning and Computer Vision, with a strong focus on building innovative algorithms for object detection, semantic segmentation, and generative modeling. My work bridges fundamental research and real-world applications, particularly in autonomous navigation and environmental monitoring.

During my Ph.D. at Complutense University of Madrid (2020–2024), advised by Professor Gonzalo Pajares and Eva Besada, I led cutting-edge projects funded by the Spanish government and the European Union, including AI-GES BLOOM-CM, SMART-BLOOMS, and AMPBAS. My contributions include:

  • Designing multi-task CNN architectures that integrate detection and segmentation for complex aquatic environments.
  • Pioneering the customization of diffusion models for synthetic data generation, coupled with LLM-guided prompt engineering to improve training pipelines.
  • Developing real-time visual perception algorithms for Autonomous Surface Vehicles (ASVs) in challenging conditions.
I have published in top-tier journals such as Expert Systems with Applications, Applied Soft Computing, and Neural Networks, with contributions on generative modeling, efficient network architectures, and robust scene understanding.

I am passionate about translating cutting-edge AI research into scalable solutions, and my current interests expand towards the synergy of Computer Vision and Large Language Models (LLMs) to tackle next-generation multimodal challenges.

I’m open to collaborations and opportunities where I can contribute to advancing AI research while building impactful, real-world applications.

Publications

Customization of the Text-to-Image Diffusion Model by Fine-Tuning for the generation of synthetic images of cyanobacterial blooms in lentic water bodies

F. Barrientos-Espillco, G. Pajares, J.A. López-Orozco, E. Besada-Portas

Expert Systems with Applications, August 2025

This paper presents a novel approach that combines DreamBooth-based fine-tuning of the Stable Diffusion XL model with LLM-driven prompts generation (LLaMa 2) to synthesize realistic images of cyanobacterial blooms and navigational obstacles in lentic water bodies.[paper]

Integration of object detection and semantic segmentation based on Convolutional Neural Networks for navigation and monitoring of cyanobacterial blooms in lentic water scenes

F. Barrientos-Espillco, M.J. Gómez-Silva, E. Besada-Portas, G. Pajares

Applied Soft Computing, September 2024

In this paper we present an architecture based on Convolutional Neural Networks (CNNs) designed to simultaneously address object detection and semantic segmentation in lentic aquatic environments.[paper]

Filter Pruning for Convolutional Neural Networks in Semantic Image Segmentation

C. I. López-González, E. Gascó, F. Barrientos-Espillco, E. Besada-Portas, G. Pajares

Neural Networks, January 2024

This paper presents innovative filter pruning and fine-tuning strategies for CNNs for use in small, resource-constrained devices, such as machine vision systems embedded in Autonomous Surface Vehicles (ASVs), where real-time computation is critical for safe navigation and contaminant detection.[paper]

Semantic segmentation based on Deep learning for the detection of Cyanobacterial Harmful Algal Blooms (CyanoHABs) using synthetic images

F. Barrientos-Espillco, E. Gascó, C.I. López-González, M.J. Gómez-Silva, G. Pajares

Applied Soft Computing, July 2023

This AI Paper proposes the detection of Cyanobacterial Harmful Algal Blooms (CyanoHABs) from the perspective of an Autonomous Surface Vehicle (ASV) using deep learning-based semantic segmentation methods with synthetic images.[paper]

Projects

  • SMART-BLOOMS (TED2021-130123B-I00) from the MCIN/AEI/10.13039/501100011033 and NextGenerationEU/PRTR programs.
  • AI-GES-BLOOM-CM: Towards an integrated system for the warning and management of cyanobacterial BLOOMs in inland waters, 2021-2024, SINERGICOS-CAM UAM-UCM.
  • AMPBAS: Automatic Monitoring of Pollutants in Dammed Waters using Biosensors and Autonamous Surface Vehicle, 2019-2021, RTI2018-098962-B-C21.

Others

I regularly collaborate as a reviewer for several journals.