Scholarlink

In this chapter, we are going to discuss Out-of-Distribution (OOD) detection, the problem of identifying “unknown” samples that do not belong to the DNN’s training distribution. We will first explain in detail what OOD detection is and why it matters. Then we introduce several representative research works characterizing two important directions, along with some recent, new explorations.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Out-of-Distribution Detection

  • 章节

  • 发表日期：

    2026年

查看全文
  • Yiran Chen,
  • Hai Li,
  • Huanrui Yang

摘要

In this chapter, we are going to discuss Out-of-Distribution (OOD) detection, the problem of identifying “unknown” samples that do not belong to the DNN’s training distribution. We will first explain in detail what OOD detection is and why it matters. Then we introduce several representative research works characterizing two important directions, along with some recent, new explorations.