AI Paper Atlas

Research map / Probabilistic and scientific modeling / Dynamical systems

Extrapolative Continuous-time Bayesian Neural Network for Fast Training-free Test-time Adaptation

Hengguan Huang, Xiangming Gu, Hao Wang, Chang Xiao, Hongfu Liu, Ye Wang

NeurIPS 2022 · Probabilistic and scientific modeling › Dynamical systems · approach: Latent-space control and planning · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
remove upper bound of temporal-domain-invariant loss (eli) (ablation)ECBNN · Streaming Rotating MNIST->USPS targetaccuracy: 53.6 -> 60.9Improved
add temporal-domain-invariant loss upper bound to ECBNN (ablation)ECBNN · Streaming Rotating MNIST->USPS out-of-domainaccuracy: 29.9 -> 38.5Improved
add PFDE loss (lde) on top of ELBO (ablation)ECBNN · Streaming Rotating MNIST->USPS targetaccuracy: 43.2 -> 53.6Improved
add PFDE loss (lde) on top of ELBO (ablation)ECBNN · Streaming Rotating MNIST->USPS out-of-domainaccuracy: 28.0 -> 29.9Improved
use ECBNN instead of ADDAADDA · Streaming Rotating MNIST->USPS targetaccuracy: 52.0 -> 60.9Improved
use ECBNN instead of ADDAADDA · Streaming Rotating MNIST->USPS out-of-domainaccuracy: 34.3 -> 38.5Improved
use ECBNN instead of Source-OnlySource-Only · Streaming Rotating MNIST->USPS targetaccuracy: 28.8 -> 60.9Improved
use ECBNN instead of EDA online adaptationEDA · OuluVS2 targetaccuracy: 70.3 -> 75.3Improved
use ECBNN instead of EDA online adaptationEDA · OuluVS2run time per batch (s): 0.23 -> 0.01Improved
use ECBNN instead of DANNDANN · OuluVS2 targetaccuracy: 71.5 -> 75.3Improved
use ECBNN instead of Source-Only on source domainSource-Only · Streaming Rotating MNIST->USPS sourceaccuracy: 97.8 -> 97.1No clear effect
increase observed history length at test timeECBNN · Growing CircleaccuracyImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy