Carlos de la Torre

Carlos de la Torre

PhD Researcher in Computer Science

University of Helsinki

Hey, I’m Carlos! 👋

I spend my time bridging Machine learning and Neuroscience, currently developing novel approaches in Brain-Computer Interfacing. My vision is having computing systems augmenting our cognitive abilities and supporting our well-being, naturally integrating with our cognition. I found home at the Cognitive Computing group led by Academy Research Fellow Tuukka Ruotsalo at the University of Helsinki.

I do some Cognitive modeling and used to spend my days slicing brains and herding cells, as I started my research in Neurobiology and Neuropharmacology.

  • Machine Learning
  • Physiological computing
  • Human-computer interaction
  • PhD Computer Science, 2021-

    University of Helsinki

  • BSc Biotechnology, 2019

    Universidad CEU San Pablo

  • BSc (MSc) Pharmacy, 2018

    Universidad CEU San Pablo

  • PCert Clinical Trials, 2017

    University of Chicago


(2022). Brain-Supervised Image Editing. CVPR.

PDF Cite Code Dataset Video

(2021). [Re] Neural Network Model of Memory Retrieval. ReScience C.

PDF Cite Code DOI

(2021). Improving Artificial Teachers by Considering How People Learn and Forget. In IUI.

PDF Cite Code Dataset Project Video DOI

(2020). Brain Relevance Feedback for Interactive Image Generation. UIST.

PDF Cite Video DOI

(2018). Endogenous pleiotrophin and midkine regulate LPS-induced glial responses. Neuroscience Letters.

Cite DOI



I use electroencephalography (EEG) recordings to implicitly control generative models’ (generative adversarial neural networks or GANs) output visual features as a model task.

🧠 What are the similarities between human cognition and machine learning models?

I investigated the similarity between images measured as GAN-learned distances and compared them to human perceptual distances (Psychophysiology, submitted).

📈 How can we use physiological recordings or medical data in user-centered machine learning methods?

I use EEG as feedback to steer the GAN and match the user’s intentions (e.g., I think of a blond and smiling person, the GAN generates a blond and smiling person UIST'20). I extend the applications to brain-based crowdsourced image generation (CHI'23, submitted) and semantic image editing (CVPR'22).

🤖 Can such data inform us to enhance or design novel machine learning algorithms?

In progress, target ICLR.

📚 Can we model human memory and help humans learn?

I have combined a cognitive model of memory (infer how the user learns and forgets) and model-based planning (when to review or show a new item) for intelligent tutoring systems IUI'21. I have also contributed to open science by replicating a connectionist model of memory, leading to the correction of the original manuscript (ReScience C).


  • Mentoring, networking, advocacy, and event planning.
  • Tech: GNU+Linux (Gentoo/Arch/Debian), FLOSS, Vim, UNIX philosophy, refurbishing computers.
  • Sports, personal growth, and music: certified coach and personal trainer. I play triad chords in a keyboard.


Machine learning

scikit-learn, PyTorch

Neurophysiological data

MNE-Python, EEG

Data science

pandas, numpy, seaborn, matplotlib



Click to zoom.