Working Student: ADA (f­­/­­m­­/­­div)

Take responsibility: You will develop a foundation model that can be easily customized and fine-tuned for various vision tasks using Infineon's dataset Data is everything: You work on a model which should be capable of accommodating images of different sizes and spatial resolutions, effectively utilizing the spectral information, and outperforming existing models for downstream tasks Expand your horizons: You will deal with Infineon's unique dataset, which differs significantly from standard image datasets Identify needs: You optimize the model to effectively utilize spectral information, potentially drawing inspiration from models like DINOv2, Mask Auto Encoder (MAE), IJEPA Experience research: You investigate multi-stage pretraining techniques to improve model performance Reliable work: You validate the model's performance through ablation studies and comparison with existing foundation models