Debiased Negative Mining Improves Out-of-distribution Detection with Pre-trained Vision-Language Models

• URL: http://arxiv.org/abs/2605.23797v1
• Date: Fri, 22 May 2026 15:57:56 GMT
• Title: Debiased Negative Mining Improves Out-of-distribution Detection with Pre-trained Vision-Language Models
• Abstract summary: Out-of-distribution (OOD) detection has emerged as a pivotal approach to enhancing the reliability of machine learning models.<n>This paper focuses on the burgeoning paradigm of post-hoc OOD detection with pre-trained vision-language models (VLMs)<n>We develop a theoretical framework for correcting the sampling bias of negatives labels by indirectly approximating the distribution of negative labels.
• Score: 30.819876174251863
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Aiming at identifying unexpected inputs from unknown classes, out-of-distribution (OOD) detection has emerged as a pivotal approach to enhancing the reliability of machine learning models. This paper focuses on the burgeoning paradigm of post-hoc OOD detection with pre-trained vision-language models (VLMs), where a popular pipeline is to detect OOD inputs by examining their affinities between ID labels and negative labels, i.e., those semantically different from ID labels. Due to the unavailability of target OOD labels, existing works predominantly rely on heuristic rules to mine negative labels from unlabeled wild corpus data. Despite the empirical success, we argue that the power of VLM-based OOD detection has yet to be fully unleashed since the notorious false negative problem is far from addressed in the literature. With this motivation, we are interested in addressing the challenge of mining true negative labels for OOD scoring. To this end, we develop a theoretical framework for correcting the sampling bias of negatives labels by indirectly approximating the distribution of negative labels. Perhaps surprisingly, we show that the debiased negative mining can be naturally converted into Monte-Carlo sampling based on ID labels and the unlabeled wild corpus data. Extensive experiments empirically manifest that our method establishes a new state-of-the-art in a variety of OOD detection setups. Code is publicly available at \href{https://github.com/60pen9/Debiased-Negative-Mining-Improves-OOD-Detection-with-Pre-trained-VLMs}{\textcolor{red}{here}}.

Related papers

• NegAS: Negative Label Guided Attention and Scoring for Out-of-Distribution Object Detection with Vision-Language Models [10.936920515353107]
  Out-of-Distribution (OOD) detection is essential for ensuring the robustness and reliability of object detection systems in safety-critical applications.<n>While prior research has mainly focused on uni-modal detectors or vision-language model (VLM) based classifiers, the potential of VLM-based object detectors in OOD scenarios remains underexplored.<n>We identify two challenges specific to VLM detectors: (i) their text-guided attention enhances foreground with ID labels but treats background uniformly, and (ii) their sigmoid-based multi-label outputs are incompatible with softmax-based OOD scores
  arXiv  Detail & Related papers  (2026-06-21T14:50:41Z)
• Activation Matters: Test-time Activated Negative Labels for OOD Detection with Vision-Language Models [22.58959194553246]
  Out-of-distribution (OOD) detection aims to identify samples that deviate from in-distribution (ID)<n>One popular pipeline addresses this by introducing negative labels distant from ID classes and detecting OOD based on their distance to these labels.<n>We propose underlineTest-time underlineActivated underlineNegative underlineLabels (TANL) by dynamically evaluating activation levels across the corpus dataset.
  arXiv  Detail & Related papers  (2026-03-26T09:53:04Z)
• NegRefine: Refining Negative Label-Based Zero-Shot OOD Detection [3.9940875711185124]
  NegRefine is a novel negative label refinement framework for zero-shot OOD detection.<n>We introduce a filtering mechanism to exclude subcategory labels and proper nouns from the negative label set.<n>We evaluate NegRefine on large-scale benchmarks, including ImageNet-1K.
  arXiv  Detail & Related papers  (2025-07-13T21:15:30Z)
• Semantic or Covariate? A Study on the Intractable Case of Out-of-Distribution Detection [70.57120710151105]
  We provide a more precise definition of the Semantic Space for the ID distribution. We also define the "Tractable OOD" setting which ensures the distinguishability of OOD and ID distributions.
  arXiv  Detail & Related papers  (2024-11-18T03:09:39Z)
• Negative Label Guided OOD Detection with Pretrained Vision-Language Models [96.67087734472912]
  Out-of-distribution (OOD) detection aims at identifying samples from unknown classes. We propose a novel post hoc OOD detection method, called NegLabel, which takes a vast number of negative labels from extensive corpus databases.
  arXiv  Detail & Related papers  (2024-03-29T09:19:52Z)
• How Does Unlabeled Data Provably Help Out-of-Distribution Detection? [63.41681272937562]
  Unlabeled in-the-wild data is non-trivial due to the heterogeneity of both in-distribution (ID) and out-of-distribution (OOD) data. This paper introduces a new learning framework SAL (Separate And Learn) that offers both strong theoretical guarantees and empirical effectiveness.
  arXiv  Detail & Related papers  (2024-02-05T20:36:33Z)
• Model-free Test Time Adaptation for Out-Of-Distribution Detection [62.49795078366206]
  We propose a Non-Parametric Test Time textbfAdaptation framework for textbfDistribution textbfDetection (abbr) abbr utilizes online test samples for model adaptation during testing, enhancing adaptability to changing data distributions. We demonstrate the effectiveness of abbr through comprehensive experiments on multiple OOD detection benchmarks.
  arXiv  Detail & Related papers  (2023-11-28T02:00:47Z)
• Dist-PU: Positive-Unlabeled Learning from a Label Distribution Perspective [89.5370481649529]
  We propose a label distribution perspective for PU learning in this paper. Motivated by this, we propose to pursue the label distribution consistency between predicted and ground-truth label distributions. Experiments on three benchmark datasets validate the effectiveness of the proposed method.
  arXiv  Detail & Related papers  (2022-12-06T07:38:29Z)
• Label Smoothed Embedding Hypothesis for Out-of-Distribution Detection [72.35532598131176]
  We propose an unsupervised method to detect OOD samples using a $k$-NN density estimate. We leverage a recent insight about label smoothing, which we call the emphLabel Smoothed Embedding Hypothesis We show that our proposal outperforms many OOD baselines and also provide new finite-sample high-probability statistical results.
  arXiv  Detail & Related papers  (2021-02-09T21:04:44Z)
• Learn what you can't learn: Regularized Ensembles for Transductive Out-of-distribution Detection [76.39067237772286]
  We show that current out-of-distribution (OOD) detection algorithms for neural networks produce unsatisfactory results in a variety of OOD detection scenarios. This paper studies how such "hard" OOD scenarios can benefit from adjusting the detection method after observing a batch of the test data. We propose a novel method that uses an artificial labeling scheme for the test data and regularization to obtain ensembles of models that produce contradictory predictions only on the OOD samples in a test batch.
  arXiv  Detail & Related papers  (2020-12-10T16:55:13Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.