Academia

SambaMixer: State of Health Prediction of Li-ion Batteries using Mamba State Space Models, IEEE Access, 2024.

José Ignacio Olalde-Verano Sascha Kirch Clara Pérez-Molina Sergio Martín

IEEE ACCESS arXiv GitHub

First application of Mamba state space models to battery health prediction, demonstrating superior performance over traditional LSTM approaches for time-series forecasting in energy systems.

RGB-D-Fusion: Image Conditioned Depth Diffusion of Humanoid Subjects, IEEE Access, 2023.

Sascha Kirch Valeria Olyunina Jan Ondřej Rafael Pagés Clara Pérez-Molina Sergio Martín

IEEE ACCESS arXiv GitHub

Novel application of diffusion models to depth estimation for volumetric video generation, developed through research collaboration with Volograms for AR/VR applications. Techniques integrated into production pipeline for single-view 3D reconstruction of human subjects.

VoloGAN: Adversarial Domain Adaptation for Synthetic Depth Data, 2022.

Sascha Kirch Rafael Pagés Sergio Arnaldo Sergio Martín

arXiv GitHub

Master's thesis work exploring GAN-based domain adaptation to bridge quality gap between high-fidelity depth maps from photogrammetry and lower-quality depth maps from mobile phone LiDAR sensors for volumetric video models.