Jun 28, 2021 � This paper focuses on the fully test-time adaptation setting, where only unlabeled data from the target distribution is required.
Nov 10, 2021 � This paper presents a method for improving DNNs performance on distribution shifts via test time adaptation based on several techniques.
Figure 1: Illustration of different losses for confidence maximization. Losses (left, shifted such that maxima of all losses are at 0) and the resulting�...
This paper proposes a novel loss that improves test-time adaptation by addressing both premature convergence and instability of entropy minimization and�...
A curated list of awesome online test-time adaptation resources. Your contributions are always welcome!
This paper focuses on the fully test-time adaptation setting, where only unlabeled data from the target distribution is required. Test-time Adaptation.
When faced with distribution shift at test time, deep neural networks often make inaccurate predictions with unreliable uncertainty estimates. While improving�...
In this paper, we propose Test-Time Self-. Training (TeST): a technique that takes as input a model trained on some source data and a novel data distribution at�...
People also ask
What is test time adaptation?
Sep 15, 2024 � This article presents a comprehensive survey of online test-time adaptation (OTTA), focusing on effectively adapting machine learning models�...
Mar 27, 2023 � Among them, Anti-Adv [479] perturbs the test input to maximize the classifier's prediction confidence, and Hedge [478] alters the test input by�...