Pith. sign in

REVIEW 10 cited by

Contrastive Training for Improved Out-of-Distribution Detection

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2007.05566 v1 pith:WDLASH5T submitted 2020-07-10 cs.LG stat.ML

Jim Winkens , Rudy Bunel , Abhijit Guha Roy , Robert Stanforth , Vivek Natarajan , Joseph R. Ledsam , Patricia MacWilliams , Pushmeet Kohli
show 5 more authors
Alan Karthikesalingam Simon Kohl Taylan Cemgil S. M. Ali Eslami Olaf Ronneberger
This is my paper · ORCID
classification cs.LGstat.ML
keywords detectioncontrastiveperformancetrainingmethodsout-of-distributionaccessapproach
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Reliable detection of out-of-distribution (OOD) inputs is increasingly understood to be a precondition for deployment of machine learning systems. This paper proposes and investigates the use of contrastive training to boost OOD detection performance. Unlike leading methods for OOD detection, our approach does not require access to examples labeled explicitly as OOD, which can be difficult to collect in practice. We show in extensive experiments that contrastive training significantly helps OOD detection performance on a number of common benchmarks. By introducing and employing the Confusion Log Probability (CLP) score, which quantifies the difficulty of the OOD detection task by capturing the similarity of inlier and outlier datasets, we show that our method especially improves performance in the `near OOD' classes -- a particularly challenging setting for previous methods.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 10 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. OpenAlex reports about 52 citations worldwide. Full citation record

  1. Outlier Synthesis via Hamiltonian Monte Carlo for Out-of-Distribution Detection

    cs.LG 2025-01 conditional novelty 7.0 of 10

    HamOS generates diverse virtual outliers via Hamiltonian Monte Carlo in hyperspherical feature space, improving OOD detection FPR95 on CIFAR-10/100 and ImageNet-1K without natural outlier data.

  2. Synthesizing Near-Boundary OOD Samples for Out-of-Distribution Detection

    cs.CV 2025-07 conditional novelty 6.0 of 10

    SynOOD generates synthetic near-boundary OOD images with MLLM-guided inpainting and energy-score gradients, then fine-tunes CLIP image and text features, reporting state-of-the-art OOD detection on ImageNet benchmarks.

  3. Balanced Hyperbolic Embeddings Are Natural Out-of-Distribution Detectors

    cs.LG 2025-06 conditional novelty 6.0 of 10

    Norm-balanced, hierarchy-aware hyperbolic prototypes as the classification head improve out-of-distribution detection across many scoring functions and benchmarks.

  4. Score Combining for Contrastive OOD Detection

    cs.LG 2025-01 conditional novelty 6.0 of 10

    A GLRT-based score-combining rule slightly improves average AUROC and detection rate over CSI/SupCSI and classical p-value combination methods in dataset-vs-dataset and leave-one-class-out OOD experiments.

  5. Function Space Diversity for Uncertainty Prediction via Repulsive Last-Layer Ensembles

    cs.LG 2024-12 conditional novelty 6.0 of 10

    A repulsive last-layer ensemble trained with function-space diversity on OOD or augmented samples gives competitive uncertainty estimates at a fraction of deep-ensemble cost.

  6. Making the Flow Glow -- Robot Perception under Severe Lighting Conditions using Normalizing Flow Gradients

    cs.CV 2024-12 conditional novelty 6.0 of 10

    Using absolute normalizing flow gradients as a pixel-level out-of-distribution score lets a robot optimize camera parameters locally, improving object detection by 60% over global-score baselines in severe lighting.

  7. DCV-ROOD Evaluation Framework: Dual Cross-Validation for Robust Out-of-Distribution Detection

    cs.LG 2025-09 conditional novelty 5.0 of 10

    DCV-ROOD is a dual cross-validation framework for OOD detection that splits ID data by stratified folds and OOD data by class groups, reproducing benchmark statistical comparisons at lower cost.

  8. Open-Set LiDAR Panoptic Segmentation Guided by Uncertainty-Aware Learning

    cs.CV 2025-06 conditional novelty 5.0 of 10

    Uncertainty-guided LiDAR panoptic segmentation (ULOPS) uses evidential learning and three uncertainty losses to segment unknown objects, outperforming prior open-set baselines on KITTI-360 and nuScenes.

  9. Self-supervised Benchmark Lottery on ImageNet: Do Marginal Improvements Translate to Improvements on Similar Datasets?

    cs.CV 2025-01 conditional novelty 5.0 of 10

    Self-supervised model rankings change substantially on ImageNet variants: DINO and Swav drop on Rendition and Sketch while MoCo and Barlow improve, so ImageNet-only benchmarking is misleading.

  10. Learning Structured Representations with Hyperbolic Embeddings

    cs.LG 2024-12 conditional novelty 5.0 of 10

    HypStructure trains image representations whose pairwise distances follow a label tree by adding a hyperbolic CPCC regularizer and a centering loss, reducing hierarchy distortion and modestly improving classification ...

Pith tools