Back to the Basics: Revisiting Out-of-Distribution Detection Baselines

07/07/2022
by   Johnson Kuan, et al.
0

We study simple methods for out-of-distribution (OOD) image detection that are compatible with any already trained classifier, relying on only its predictions or learned representations. Evaluating the OOD detection performance of various methods when utilized with ResNet-50 and Swin Transformer models, we find methods that solely consider the model's predictions can be easily outperformed by also considering the learned representations. Based on our analysis, we advocate for a dead-simple approach that has been neglected in other studies: simply flag as OOD images whose average distance to their K nearest neighbors is large (in the representation space of an image classifier trained on the in-distribution data).

READ FULL TEXT

page 1

page 2

page 3

page 4

research
09/04/2022

A Novel Nearest Neighbors Algorithm Based on Power Muirhead Mean

This study aimed to propose a novel classifier based on K-Nearest Neighb...
research
06/01/2023

StableRep: Synthetic Images from Text-to-Image Models Make Strong Visual Representation Learners

We investigate the potential of learning visual representations using sy...
research
11/12/2022

Far Away in the Deep Space: Nearest-Neighbor-Based Dense Out-of-Distribution Detection

The key to out-of-distribution detection is density estimation of the in...
research
06/02/2023

A Simple yet Effective Self-Debiasing Framework for Transformer Models

Current Transformer-based natural language understanding (NLU) models he...
research
03/13/2018

Deep k-Nearest Neighbors: Towards Confident, Interpretable and Robust Deep Learning

Deep neural networks (DNNs) enable innovative applications of machine le...
research
02/08/2023

Robustness to Spurious Correlations Improves Semantic Out-of-Distribution Detection

Methods which utilize the outputs or feature representations of predicti...
research
10/27/2022

On Out-of-Distribution Detection for Audio with Deep Nearest Neighbors

Out-of-distribution (OOD) detection is concerned with identifying data p...

Please sign up or login with your details

Forgot password? Click here to reset