Few-Shot Out of Domain Intent Detection with Covariance Corrected Mahalanobis Distance

2026-09-01

AI Summary

本文针对少样本场景下领域外意图检测问题，通过提出一种协方差校正的马氏距离方法来改进检测效果。

Abstract

Conversational agents like chatbots and voice assistants are trained to understand and respond to user intents. On encountering an utterance with an intent different from the ones they have been trained on, these agents are expected to classify the intent as `unknown' or `out of domain'. This problem is known as out of domain (OOD) intent detection. Podolskiy et al. (2021), showed that Mahalanobis distance can be used effectively for identifying OOD intents, outperforming competing approaches. However, their method fails to outperform the baselines in the practically important few-shot setting. In this paper we analyze the reason for low performance and propose a covariance corrected Mahalanobis distance for detecting out-of-domain intents.

Problem

• Few-Shot
• Out of Domain
• Intent Detection

Innovation

• Covariance Corrected Mahalanobis Distance
• Few-Shot Learning
• Out of Domain Intent Detection