摘要
半监督指代表达分割旨在有限标注下实现精确像素级语言定位，但常受限于监督不足及伪标签不可靠。本文提出
AI 推荐理由
论文提出自进化框架，通过强化学习循环优化伪标签与模型，核心机制符合自我进化定义。
研究机构
School of Computer Science and Engineering, Nanjing University of Science and Technology
School of Information Engineering, Huzhou Normal University
School of Artificial Intelligence, Nanjing University of Posts and Telecommunications
National Key Laboratory of Tibetan Language Intelligence
Beijing Normal University
论文信息