U-Net Noise Robustness Study

Job ID: 40152471

Budget: €250 – €750 EUR

I need help designing an experiment to study the impact of Pelt/Okumoto (Gaussian) noise and label error on U-Net performance. The focus is on measuring robustness to noise. Dimensions to be tested are label adaption vs header adaption

Requirements:
- Experiment design expertise
- Understanding of U-Net and noise effects (Pelt/ Okumoto)
- Ability to measure and analyze robustness metrics

Related categories: Data Processing Machine Learning (ML) Statistical Analysis Data Science Neural Networks Signal Processing Data Visualization Data Analysis Deep Learning Model Testing & Optimization