Revisit Overconfidence for OOD Detection: Reassigned Contrastive Learning with Adaptive Class-dependent Threshold
Detecting Out-of-Domain (OOD) or unknown intents from user queries is essential in a task-oriented dialog system. A key challenge of OOD detection is the overconfidence of neural models. In this paper, we comprehensively analyze overconfidence and classify it into two perspectives: over-confident OOD and in-domain (IND). Then according to intrinsic reasons, we respectively propose a novel reassigned contrastive learning (RCL) to discriminate IND intents for over-confident OOD and an adaptive class-dependent local threshold mechanism to separate similar IND and OOD intents for over-confident IND. Experiments and analyses show the effectiveness of our proposed method for both aspects of overconfidence issues.
Code (1)
Tasks
Contrastive LearningOut of Distribution (OOD) DetectionSimilar Papers 제목 키워드 기반
eMargin: Revisiting Contrastive Learning with Margin-Based Separation
We revisit previous contrastive learning frameworks to investigate the effect of introducing an adaptive margin into the contrastive loss function for time series representation learning. Specifically, we explore whether…
Representation LearningContrastive LearningSimplification Is All You Need against Out-of-Distribution Overconfidence
Deep neural networks (DNNs) often exhibit out-of-distribution (OOD) overconfidence, producing overly confident predictions on OOD samples. We attribute this issue to the inherent over-complexity of DNNs and investiga…
AllAttributeKnowledge DistillationMitigating Overconfidence in Out-of-Distribution Detection by Capturing Extreme Activations
Detecting out-of-distribution (OOD) instances is crucial for the reliable deployment of machine learning models in real-world scenarios. OOD inputs are commonly expected to cause a more uncertain prediction in the primar…
Out-of-Distribution DetectionTowards Efficient and General-Purpose Few-Shot Misclassification Detection for Vision-Language Models
Reliable prediction by classifiers is crucial for their deployment in high security and dynamically changing situations. However, modern neural networks often exhibit overconfidence for misclassified predictions, highlig…
Prompt LearningRevisiting Contrastive Methods for Unsupervised Learning of Visual Representations
Contrastive self-supervised learning has outperformed supervised pretraining on many downstream tasks like segmentation and object detection. However, current methods are still primarily applied to curated datasets like …
Instance Segmentationobject-detectionObject DetectionRetrieval+3