You can have your ensemble and run it too - Deep Ensembles Spread Over Time
Paper in proceeding, 2023

Ensembles of independently trained deep neural networks yield uncertainty estimates that rival Bayesian networks in performance. They also offer sizable improvements in terms of predictive performance over single models. However, deep ensembles are not commonly used in environments with limited computational budget - such as autonomous driving - since the complexity grows linearly with the number of ensemble members. An important observation that can be made for robotics applications, such as autonomous driving, is that data is typically sequential. For instance, when an object is to be recognized, an autonomous vehicle typically observes a sequence of images, rather than a single image. This raises the question, could the deep ensemble be spread over time?In this work, we propose and analyze Deep Ensembles Spread Over Time (DESOT). The idea is to apply only a single ensemble member to each data point in the sequence, and fuse the predictions over a sequence of data points. We implement and experiment with DESOT for traffic sign classification, where sequences of tracked image patches are to be classified. We find that DESOT obtains the benefits of deep ensembles, in terms of predictive and uncertainty estimation performance, while avoiding the added computational cost. Moreover, DESOT is simple to implement and does not require sequences during training. Finally, we find that DESOT, like deep ensembles, outperform single models for out-of-distribution detection.

traffic sign recognition

uncertainty estimation

out of distribution detection

ensemble

Author

Isak Meding

Zenseact AB

Alexander Bodin

Zenseact AB

Adam Tonderski

Lund University

Zenseact AB

Joakim Johnander

Zenseact AB

Linköping University

Christoffer Petersson

Chalmers, Mathematical Sciences, Algebra and geometry

Zenseact AB

Lennart Svensson

Chalmers, Electrical Engineering, Signal Processing and Biomedical Engineering

Proceedings - 2023 IEEE/CVF International Conference on Computer Vision Workshops, ICCVW 2023

4022-4031
9798350307443 (ISBN)

2023 IEEE/CVF International Conference on Computer Vision Workshops, ICCVW 2023
Paris, France,

Subject Categories

Information Science

Computer Science

Computer Vision and Robotics (Autonomous Systems)

DOI

10.1109/ICCVW60793.2023.00434

More information

Latest update

2/2/2024 1