Deep Ensembles (DE) are a prominent approach for achieving excellent performance on key metrics such as accuracy, calibration, uncertainty estimation, and out-of-distribution detection. However, hardware limitations of real-world systems constrain to smaller ensembles and lower-capacity networks, significantly deteriorating their performance and properties. We introduce Packed-Ensembles (PE), a strategy to design and train lightweight structured ensembles by carefully modulating the dimension of their encoding space. We leverage grouped convolutions to parallelize the ensemble into a single shared backbone and forward pass to improve training and inference speeds. PE is designed to operate within the memory limits of a standard neural network. Our extensive research indicates that PE accurately preserves the properties of DE, such as diversity, and performs equally well in terms of accuracy, calibration, out-of-distribution detection, and robustness to distribution shift. We make our code available at https://github.com/ENSTA-U2IS/torch-uncertainty.
@article{arxiv.2210.09184,
title = {Packed-Ensembles for Efficient Uncertainty Estimation},
author = {Olivier Laurent and Adrien Lafage and Enzo Tartaglione and Geoffrey Daniel and Jean-Marc Martinez and Andrei Bursuc and Gianni Franchi},
journal= {arXiv preprint arXiv:2210.09184},
year = {2025}
}
Comments
Published as a conference paper at ICLR 2023 (notable 25%)