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

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

12 Nov 2022arXiv:2211.06660archive 2025-07-28

Silvio Galesso, Max Argus, Thomas Brox

The key to out-of-distribution detection is density estimation of the in-distribution data or of its feature representations. This is particularly challenging for dense anomaly detection in domains where the in-distribution data has a complex underlying structure. Nearest-Neighbors approaches have been shown to work well in object-centric data domains, such as industrial inspection and image classification. In this paper, we show that nearest-neighbor approaches also yield state-of-the-art results on dense novelty detection in complex driving scenes when working with an appropriate feature representation. In particular, we find that transformer-based architectures produce representations that yield much better similarity metrics for the task. We identify the multi-head structure of these models as one of the reasons, and demonstrate a way to transfer some of the improvements to CNNs. Ultimately, the approach is simple and non-invasive, i.e., it does not affect the primary segmentation performance, refrains from training on examples of anomalies, and achieves state-of-the-art results on RoadAnomaly, StreetHazards, and SegmentMeIfYouCan-Anomaly.

Paper PDF Code

Code

silviogalesso/dense-ood-knns officialmentioned in papermentioned on GitHubpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Anomaly Detection Density Estimation Image Classification Novelty Detection Out-of-Distribution Detection Semantic Segmentation image-classification

Results from the paper archive 2025-07-28

Task	Dataset	Model	Metric	Value	Rank at snapshot	Leaderboard	Report
Anomaly Detection	Fishyscapes L&F	cDNP+OE	AP	69.8	#1 of 18	Archive leaderboard	report
Anomaly Detection	Fishyscapes L&F	cDNP+OE	FPR95	7.5	#1 of 18	Archive leaderboard	report
Anomaly Detection	Fishyscapes L&F	cDNP	AP	62.2	#3 of 18	Archive leaderboard	report
Anomaly Detection	Fishyscapes L&F	cDNP	FPR95	8.9	#3 of 18	Archive leaderboard	report
Anomaly Detection	Road Anomaly	cDNP	AP	85.6	#4 of 10	Archive leaderboard	report
Anomaly Detection	Road Anomaly	cDNP	FPR95	9.8	#4 of 10	Archive leaderboard	report
Out-of-Distribution Detection	ADE-OoD	cDNP	AP	62.35	#3 of 4	Archive leaderboard	report
Out-of-Distribution Detection	ADE-OoD	cDNP	FPR@95	39.20	#3 of 4	Archive leaderboard	report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections