@inproceedings{,
    author = {Baumann, Anton; Roßberg, Thomas; Schmitt, Michael},
    title = {Probabilistic MIMO U-Net: Efficient and Accurate Uncertainty Estimation for Pixel-Wise Regression},
    editor = {},
    booktitle = {2023 IEEE/CVF International Conference on Computer Vision Workshops (ICCVW)},
    series = {},
    journal = {},
    address = {Piscataway, NJ},
    publisher = {IEEE},
    edition = {},
    year = {2023},
    isbn = {979-8-3503-0744-3},
    volume = {},
    number =  {},
    pages = {4500-4508},
    url = {},
    doi = {10.1109/ICCVW60793.2023.00484},
    keywords = {},
    abstract = {Uncertainty estimation in machine learning is paramount for enhancing the reliability and interpretability of predictive models, especially in high-stakes real-world scenarios. Despite the availability of numerous methods, they often pose a trade-off between the quality of uncertainty estimation and computational efficiency. Addressing this challenge, we present an adaptation of the Multiple-Input Multiple-Output (MIMO) framework – an approach exploiting the overparameterization of deep neural networks – for pixel-wise regression tasks. Our MIMO variant expands the applicability of the approach from simple image classification to broader computer vision domains. For that purpose, we adapted the U-Net architecture to train multiple subnetworks within a single model, harnessing the overparameterization in deep neural networks. Additionally, we introduce a novel procedure for synchronizing subnetwork performance within the MIMO framework. Our comprehensive evaluations of the resulting MIMO U-Net on two orthogonal datasets demonstrate comparable accuracy to existing models, superior calibration on in-distribution data, robust out-of-distribution detection capabilities, and considerable improvements in parameter size and inference time. Code available at github.com/antonbaumann/MIMO-Unet.},
    note = {},
    institution = {Universität der Bundeswehr München, Fakultät für Luft- und Raumfahrttechnik, LRT 9 - Institut für Raumfahrttechnik und Weltraumnutzung, Professur: Schmitt, Michael},
    
    
}