Ilia Semenkov

PhD, Research Scientist (AI/ML)

isemenkov@duck.com | LinkedIn | GitHub | Google Scholar | Website

Summary

AI/ML research scientist with end-to-end ownership across applied and methodological projects, from problem framing and reproducible experimentation to publication and open-source release. Recent work spans neuroimaging and biosignals, with parallel contributions in multimodal learning, interpretable models, and computer vision benchmarking.

Experience

Research Fellow | Artificial Intelligence Research Institute

Dec 2021 - Present | Moscow, Russia

  • Drive AI/ML methodology within research projects, from problem formulation and model design to baselines, evaluation protocols, and ablation strategy.
  • Build end-to-end research pipelines and run publication-grade experiments with strong emphasis on reproducibility, robustness, and analysis quality.
  • Co-author papers through the full research cycle, including implementation, additional experiments, error analysis, and revision rounds.

Research Fellow | HSE University

Oct 2023 - Present | Moscow, Russia

  • Translate scientific questions from interdisciplinary collaborators into concrete ML tasks, data formulations, and validation setups.
  • Provide technical ML leadership across collaborative projects through architecture reviews, experiment design, and quality control of results.
  • Supervise students and junior researchers across theses, internships, and project work, helping scope problems and review implementations and findings.

Scientific Advisor | ITMO University

Jan 2024 - Present | Moscow, Russia

  • Advise MSc research in AI/ML by shaping research questions, methodology, and evaluation plans.
  • Review experiments, results, and thesis narratives to bring projects to a coherent and technically sound final form.

Earlier Experience

  • Research Intern | HSE University | Apr 2020 - Jun 2021 | Moscow, Russia
  • Teaching Assistant | HSE University | Jan 2017 - Jun 2021 | Moscow, Russia
  • Research Intern | Institute for Problems of Information Transmission | Feb 2020 - Apr 2021 | Moscow, Russia

Selected Projects

Compact Interpretable Neural Decoders

Slim decoding architectures that expose spatial and temporal structure instead of hiding it inside opaque representations.

Safer Brain Surgery Speech Mapping

Stimulation-free mapping of speech-related cortex from ECoG as a safer complement to direct electrical stimulation.

Real-Time Brain Signal Tracking

Millisecond-scale tracking of neural rhythm envelope and phase for closed-loop EEG applications.

Camera Vision Benchmarks at Scale

Large-scale datasets and challenge protocols for more reliable evaluation in computational color constancy.

Selected Publications and Outputs

Real-time low latency estimation of brain rhythms with deep neural networks

Journal of Neural Engineering, 2023

Representational dissimilarity component analysis (ReDisCA)

NeuroImage, 2024

Towards stimulation-free automatic electrocorticographic speech mapping in neurosurgery patients

Journal of Neural Engineering, 2025

The Cube++ Illumination Estimation Dataset

IEEE Access, 2020

SIGNAL: Dataset for Semantic and Inferred Grammar Neurological Analysis of Language

Scientific Data, 2025

Education

Ph.D. in Computer Science | HSE University | 2021 - 2025

Thesis: "Neural network-based methods for decoding multimodal neuroimaging data"

M.S. in Computer Science | HSE University | 2019 - 2021

B.S. in Public Administration | HSE University | 2015 - 2019

Skills

Expertise

Deep learning | Representation learning | Multimodal learning | Neuroimaging (EEG, MEG, fMRI, ECoG) | Interpretable ML | Computer vision

Tech

Python | PyTorch | Git | Linux | Slurm

Languages

English (fluent) | Russian (native)