PhD candidate · University of Murcia

Adrián Cánovas Rodríguez

I work on artificial intelligence — deep learning on heterogeneous, real-world data, distributed and federated learning, and the security and privacy of AI models. I care about AI that is reliable, trustworthy and works outside the lab.

Download CV Get in touch
Portrait of Adrián Cánovas Rodríguez
Spain
  • Research stay IBM Research Europe Dublin · May–Aug 2026
  • PhD Computer Engineering International Mention · 2024–
  • Teaching Associate Professor University of Murcia · 2024–
01

About

I am a PhD candidate in Computer Engineering at the University of Murcia (Department of Information and Communication Engineering), working on Knowledge Acquisition through Artificial Intelligence Techniques for Sustainable Environments. Alongside my research I am an Associate Professor in the same university, teaching in the Computer Engineering programme.

Before the PhD I spent several years at IMIDA developing deep learning models on sensor, satellite and weather data. Today my research spans multi-source learning, distributed and federated AI, and the security and privacy of machine learning models — in 2026 I joined IBM Research Europe in Dublin as a research intern, studying vulnerabilities in large language models.

Research interests

  • Distributed & federated AI

    Learning across devices and institutions without centralising data, from AIoT at the edge to the cloud.

  • Security & privacy in ML

    Adversarial attacks, data leakage and privacy-preserving training, including LLMs.

  • Multi-source learning

    Deep learning and computer vision on time series and imagery from satellites, sensors and weather data.

  • AI in the real world

    Robust models under domain shift, deployed on edge and cloud infrastructure.

02

Publications

Full list on ORCID.

  1. 2026

    AIoT-Driven Pest Monitoring Approach for Real-Time Tuta absoluta Detection and Population Prediction in a Controlled Greenhouse Environment

    M. Bibi, S. Z. Khan, A. Cánovas-Rodríguez, M. Á. González-Illán, M. F. García-Cruz, P. J. Fernández-Campillo, A. F. Skarmeta, M. Á. Zamora-Izquierdo

    Preprints

    Preprint
  2. 2026
  3. 2026

    Quantifying and Mitigating Domain Shift in Peach Leaf Damage Classification: Attention Mechanisms and Fine-Tuning Strategies

    A. Cánovas-Rodríguez, M. A. González-Illán, M. F. García-Cruz, P. Nortes Tortosa, J. S. Rubio-Asensio, M. A. Zamora-Izquierdo, J. A. Martínez-Navarro, A. F. Skarmeta

    arXiv:2606.02045

    Paper Code
  4. 2025

    Chlorophyll-a Mapping and Prediction in the Mar Menor Lagoon Using C2RCC-Processed Sentinel-2 Imagery

    A. Martínez-Ibarra, A. González-Vidal, A. Cánovas-Rodríguez, A. F. Skarmeta

    arXiv:2510.09736

    Paper Data
03

Projects

More on GitHub.

turbidity_satellite_estimation

Surface turbidity estimation in the Mar Menor lagoon from Planet SuperDove imagery, calibrated against in-situ CTD measurements, with leakage-aware CatBoost and gradient boosting models.

  • Jupyter
  • PlanetScope
  • Water quality

soil-moisture-crop-propagation

Soil moisture at five depths from Sentinel-2 indices and optimally-lagged meteorological features. LSTM, CNN and hybrid models under a leak-free temporal evaluation protocol.

  • Python
  • Sentinel-2
  • Time series

peach-leaf-cbam

Attention-enhanced CNNs (CBAM) and transfer learning for peach leaf damage classification under domain shift, plus an annotated benchmark.

  • Jupyter
  • PyTorch
  • Computer vision

evaluation_desertification

Desertification assessment and forecasting from a decade of Sentinel-2 time series (2015–2025), with a Desertification Degree Index and BiLSTM / ConvLSTM forecasting.

  • Jupyter
  • Sentinel-2
  • ConvLSTM

ForestAnnotationRefinement

Effect of manual annotation refinement on instance segmentation accuracy in UAV-LiDAR-based forest inventories.

  • Jupyter
  • Segmentation
  • UAV-LiDAR

sat-img-download

Downloads, cloud-masks and crops imagery from the Copernicus Data Space Ecosystem — the data layer behind most of the projects above.

  • Jupyter
  • Copernicus
  • Tooling
04

Experience

  1. May – Aug 2026

    Research Intern · IBM Research Europe, Dublin

    Security and privacy vulnerabilities in large language models: adversarial attacks, prompt injection, data leakage and privacy-preserving fine-tuning.

  2. 2025 – present

    Researcher · University of Murcia

    Deep learning on large-scale sensor and geospatial data, privacy-preserving distributed AI, and applications to environmental monitoring and agriculture.

  3. 2024 – present

    Associate Professor · University of Murcia

    Undergraduate teaching in the Computer Engineering programme.

  4. 2023 – 2025

    Research Assistant · IMIDA

    AI pipelines for crop monitoring, soil analysis and agronomic decision support, built together with domain experts.

  5. 2021 – 2023

    Junior Software Engineer · TICARUM, University of Murcia

    Enterprise software to automate university administrative processes, across the full development lifecycle.

  6. 2018 – 2021

    Research Intern · IMIDA

    Meteorological data pipelines (SIAM, AEMET) and ML models for evapotranspiration and crop condition estimation.

Education

  • 2024 – present

    Ph.D. in Computer Engineering

    University of Murcia · International Mention

  • 2022 – 2024

    M.Sc. in New Technologies in Computer Engineering

    University of Murcia · Thesis on soil moisture and crop monitoring with satellite imagery and deep learning

  • 2016 – 2021

    B.Sc. in Computer Engineering

    University of Murcia · Thesis on reference evapotranspiration estimation with machine learning

Toolbox

Machine learning
PythonPyTorchTensorFlowscikit-learnHugging FaceOpenCV
Data & geospatial
Sentinel-2PlanetScopeCopernicusESA SNAPGoogle Earth EngineQGIS
Infrastructure
LinuxDockerKubernetesGitSQLLaTeX

Curriculum Vitae

Education, experience, publications and skills in a single PDF.

Download CV adriancr@um.es