Skip to content
View andropar's full-sized avatar
✨
✨
13 followers · 1 following

Achievements

Achievement: QuickdrawAchievement: Pair Extraordinairex4Achievement: Pull Sharkx3Achievement: YOLO

Achievements

Achievement: QuickdrawAchievement: Pair Extraordinairex4Achievement: Pull Sharkx3Achievement: YOLO

Block or report andropar

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
andropar/README.md

Johannes Roth

I'm a computer scientist and ML engineer with a background in business information systems, applied data science, and computational neuroscience.

Right now, I'm finishing a PhD in Computer Science / Machine Learning at the Max Planck Institute in Leipzig, where I build large-scale data and ML infrastructure for NeuroAI and visual neuroscience. Before that, I worked on production-facing ML services, recommender systems, image-processing pipelines, and product data workflows.

My current work asks how we can choose better images for vision and fMRI experiments: selecting from millions of natural photographs, comparing model and brain representations, and building research software that is meant to be reused rather than run once for a paper.

Selected Work

  • How to sample the world for understanding the visual system: code and experiments for a CCN 2025 paper on sampling diverse natural images for visual neuroscience.
  • LAION-fMRI: deeply sampled 7T fMRI responses to 25k natural images for NeuroAI and visual neuroscience.
  • ReLAION-2B Natural: naturalness scores for 2.1B images, identifying photographic images useful for vision research.
  • thingsvision: Python toolbox for extracting and comparing representations from 100+ vision models.
  • slurmboard: lightweight dashboard for monitoring Slurm jobs, logs, and resource usage on HPC clusters.

Links

Pinned Loading

  1. how-to-sample how-to-sample Public

    Code and experiments for CCN 2025 paper on sampling diverse natural images for visual neuroscience.

    Jupyter Notebook 3

  2. pysearchlight pysearchlight Public

    Customisable searchlight analysis for Python

    Python 1

  3. slurmboard slurmboard Public

    Lightweight dashboard for monitoring Slurm jobs, logs, and resource usage on HPC clusters.

    JavaScript 1

  4. ViCCo-Group/thingsvision ViCCo-Group/thingsvision Public

    Python package for extracting representations from state-of-the-art computer vision models

    Python 178 24

  5. conwell_replication conwell_replication Public

    LAION-fMRI replication of Conwell et al. (2024) model-brain alignment analyses with DeepNSD-style feature extraction.

    Python

  6. ViCCo-Group/LAION-fMRI ViCCo-Group/LAION-fMRI Public

    The LAION-fMRI data documentation

    Python 7 3