TINS: Test-time ID-prototype-separated Negative Semantics Learning for OOD Detection

2026-05-11 · Yifeng Yang, Jubo Feng, Jing Xu, Xinbing Wang, Qinying Gu, Nanyang Ye arxiv

Vision-language models enable OOD detection by comparing image alignment with ID labels and negative semantics. Existing negative-label-based methods mainly rely on static negative labels constructed before inference, limiting their ability to cover diverse and evolving OOD concepts. Although test-time expansion provides a natural solution, naively learning negative semantics from potential OOD samples may introduce hard ID contamination. To address this issue, we propose a \textbf{T}est-time \textbf{I}D-prototype-separated \textbf{N}egative \textbf{S}emantics learning method, termed \textbf{TINS}. TINS learns sample-specific negative text embeddings via image-to-text modality inversion and introduces ID-prototype-separated regularization to keep them separated from ID semantics. To further stabilize negative semantics expansion, TINS employs group-wise aggregation scoring and a buffer update strategy. Extensive experiments across Four-OOD, OpenOOD, Temporal-shift, and Various ID settings show consistent improvements over strong baselines. Notably, on the Four-OOD benchmark with ImageNet-1K as ID, TINS reduces the average FPR95 from 14.04\% to 6.72\%. Our code is available at https://github.com/zxk1212/tins.

📄 PDF Abstract BibTeX arXiv:2605.10756

Code (0)

등록된 구현이 없습니다.

FastInst: A Simple Query-Based Model for Real-Time Instance Segmentation

2023-03-15 · CVPR 2023 1 · Junjie He, Pengyu Li, Yifeng Geng, Xuansong Xie

Recent attention in instance segmentation has focused on query-based models. Despite being non-maximum suppression (NMS)-free and end-to-end, the superiority of these models on high-accuracy real-time benchmarks has not …

Decoder Instance Segmentation Real-time Instance Segmentation Segmentation+1

Gradient Alignment with Prototype Feature for Fully Test-time Adaptation

2024-02-14 · Juhyeon Shin, Jonghyun Lee, Saehyung Lee, MinJun Park 외

In context of Test-time Adaptation(TTA), we propose a regularizer, dubbed Gradient Alignment with Prototype feature (GAP), which alleviates the inappropriate guidance from entropy minimization loss from misclassified pse…

Pseudo Label Test-time Adaptation

Class-Aware Prototype Learning with Negative Contrast for Test-Time Adaptation of Vision-Language Models

2025-10-22 · Xiaozhen Qiao, Jingkai Zhao, Yuqiu Jiang, Xianda Guo 외 arxiv

Vision-Language Models (VLMs) demonstrate impressive zero-shot generalization through large-scale image-text pretraining, yet their performance can drop once the deployment distribution diverges from the training distrib…

Zero-shot Generalization Test-time Adaptation Contrastive Learning

Regret Analysis of the Finite-Horizon Gittins Index Strategy for Multi-Armed Bandits

2015-11-18 · Tor Lattimore

I analyse the frequentist regret of the famous Gittins index strategy for multi-armed bandits with Gaussian noise and a finite horizon. Remarkably it turns out that this approach leads to finite-time regret guarantees co…

Multi-Armed Bandits Thompson Sampling

Beyond Static Anchors: Bounded Prototype Conditioning for Language-Free Medical Anomaly Detection

2026-08-01 · Yibo Wan, Jinyu Cai, Seekiong-Ng arxiv

Medical anomaly detection identifies abnormal images and localizes lesions under scarce supervision while generalizing across organs and modalities. Existing CLIP-based methods reduce annotation requirements through visi…

Anomaly Detection