Evaluating Scalable Bayesian Deep Learning Methods for Robust Computer Vision

Gustafsson, Fredrik K.Uppsala universitet,Avdelningen för systemteknik,Artificiell intelligens (författare)

While deep neural networks have become the go-to approach in computer vision, the vast majority of these models fail to properly capture the uncertainty inherent in their predictions. Estimating this predictive uncertainty can be crucial, for example in automotive applications. In Bayesian deep learning, predictive uncertainty is commonly decomposed into the distinct types of aleatoric and epistemic uncertainty. The former can be estimated by letting a neural network output the parameters of a certain probability distribution. Epistemic uncertainty estimation is a more challenging problem, and while different scalable methods recently have emerged, no extensive comparison has been performed in a real-world setting. We therefore accept this task and propose a comprehensive evaluation framework for scalable epistemic uncertainty estimation methods in deep learning. Our proposed framework is specifically designed to test the robustness required in real-world computer vision applications. We also apply this framework to provide the first properly extensive and conclusive comparison of the two current state-of-the-art scalable methods: ensembling and MC-dropout. Our comparison demonstrates that ensembling consistently provides more reliable and practically useful uncertainty estimates. Code is available at https://github.com/fregu856/evaluating_bdl.

Danelljan, MartinSwiss Fed Inst Technol, Comp Vis Lab, Zurich, Switzerland. (författare)
Schön, Thomas B.,Professor,1977-Uppsala universitet,Avdelningen för systemteknik,Artificiell intelligens(Swepub:uu)thosc112 (författare)
Uppsala universitetAvdelningen för systemteknik (creator_code:org_t)

Ingår i:2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW 2020): IEEE Computer Society, s. 1289-12989781728193601

Av författaren/redakt...: Gustafsson, Fred ...; Danelljan, Marti ...; Schön, Thomas B. ...

Om ämnet

NATURVETENSKAP: NATURVETENSKAP; och Data och informa ...; och Datorseende och ...

Kungliga biblioteket hanterar dina personuppgifter i enlighet med EU:s dataskyddsförordning (2018), GDPR. Läs mer om hur det funkar här.
Så här hanterar KB dina uppgifter vid användning av denna tjänst.