HELLO, MY NAME IS

Gustavo Scheidt

Master Student in
Computer Vision & AI

gustavomeloscheidt@gmail.com
+33 07 53 65 45 94
Foto
About Me

I’m a Master’s student in Computer Vision and Artificial Intelligence at Institut Polytechnique de Paris, passionate about applying deep learning to healthcare and 3D computer vision. My goal is to bridge advanced AI research with real-world impact through innovation and collaboration.

Projects

VM-UNet 3D

VM-UNet 3D

3D adaptation of VM-UNet with Mamba blocks for brain MRI segmentation (BraTS2021), integrated into nnU-Net v2.

View on GitHub
Endless Runner

Endless Runner (Unity)

Complete 3D runner game with multiple stages, custom UI/UX, and full flow (menu → stage select → gameplay → game over).

View on GitHub
Narrow AI for Tourism

Narrow AI — Tourism Assistant

Specialized NLP for touristic applications with multilingual intent detection and entity extraction.

Demo / Repo
Gaussian Splatting & Volumetric Capture

Gaussian Splatting & Volumetric Capture

Real-time 3D/4D radiance field reconstruction using SfM and multi-view alignment. Implemented a full Gaussian Splatting pipeline with opacity and densification control for dynamic volumetric scenes.

Details
Iniciação Científica — Biomedical IoT

Iniciação Científica — Biomedical IoT

Developed a non-invasive wearable prototype for early cardiac anomaly detection, integrating sensor data acquisition, real-time monitoring, and Machine Learning algorithms for anomalies detection using Python.

Details
Hackathon Project

Hackathon Project

Non-invasive diabetes monitoring platform using AI and cloud microservices; built end-to-end (API, frontend, data pipeline).

Case / Repo

My Experience

2024–2025
ARTEMIS Lab — Télécom SudParis
Palaiseau, FR

Research Internship — Medical Image Segmentation

Student Researcher — Medical Imaging

  • Built a 3D VM-UNet within nnU-Net v2 on BraTS2021; ran ablations with Mamba blocks.
  • Set up a reproducible pipeline (preprocessing, augmentation, experiment tracking; metrics Dice/HD95).
  • Key project: VM-UNet 3D (see Projects).
2023–2024
UNICAMP — Research
Campinas, BR

Scientific Initiation — Biomedical IoT

Wearable for the non-invasive detection of cardiac anomalies.

  • Acquisition and signal analysis in Python with ML baselines.

Awards

2024–2025

Projet Cassiopée — Winner (IP Paris)

Achieved the 2nd highest score among 100+ teams at IP Paris. The project evaluated and developed Mamba-based neural architectures for 3D brain tumor segmentation within the nnU-Net framework. details

2024

Brazilian Stage Winner — Health Systems Innovation Lab (Harvard)

Non-invasive diabetes detection MVP built in 48h; role: ML/Backend; pitched to jury.

Skills

AI / ML
PyTorch nnU-Net v2 OpenCV scikit-learn NumPy Pandas LLMs Transformers Mamba
Programming
PythonC C++VHDL
Modeling & Simulation
MATLABSimulink
CAD / EDA
HFSSAutoCAD
Hardware & Platforms
FPGAArduino
Game Dev
UnityC#
Web / CMS
WordPress

Languages

  • Portuguese — Native
  • English — C1
  • French — B2
  • Spanish — B1
  • German — A1