PixOOD: Pixel-Level Out-of-Distribution Detection

Autor:	Vojíř, Tomáš, Šochman, Jan, Matas, Jiří
Rok vydání:	2024
Předmět:	Computer Science - Computer Vision and Pattern Recognition
Druh dokumentu:	Working Paper
Popis:	We propose a dense image prediction out-of-distribution detection algorithm, called PixOOD, which does not require training on samples of anomalous data and is not designed for a specific application which avoids traditional training biases. In order to model the complex intra-class variability of the in-distribution data at the pixel level, we propose an online data condensation algorithm which is more robust than standard K-means and is easily trainable through SGD. We evaluate PixOOD on a wide range of problems. It achieved state-of-the-art results on four out of seven datasets, while being competitive on the rest. The source code is available at https://github.com/vojirt/PixOOD. Comment: published at ECCV2024, table 1,2 improved results for the PixOOD variants thanks to fixing bug in normalization of input image
Databáze:	arXiv
Externí odkaz:	http://arxiv.org/abs/2405.19882 Zobrazit plný text záznamu View this record from Arxiv