SignTrace: Describe a Sign, Find the Word
SignTrace: Describe a Sign, Find the Word
SignTrace：描述手势，查找词义
Identifying an unfamiliar sign is difficult when a learner remembers its movement but does not know its meaning or formal feature codes. SignTrace addresses this longstanding reverse-lookup problem through natural-language access to a Chinese sign-language dictionary.
当学习者记得某个手势的动作，却不知道其含义或正式特征代码时，识别该陌生手势会非常困难。SignTrace 通过自然语言访问中国手语词典，解决了这一长期存在的反向查询难题。
The system integrates LLM-based dictionary enrichment, action extraction, dictionary-style rewriting, seven-channel retrieval, and candidate reranking over 6,699 entries. It has been deployed for user trials and has received positive informal feedback.
该系统集成了基于大语言模型（LLM）的词典扩充、动作提取、词典风格重写、七通道检索以及针对 6,699 个词条的候选重排序功能。目前该系统已投入用户试用，并获得了积极的非正式反馈。
Evaluation on a dictionary-derived benchmark of 500 movement-description queries yields 94.0% Hit@1, 97.4% Hit@9, and a mean reciprocal rank of 0.9540. Reranking increases Hit@1 from 71.8% to 94.0%, while component analyses show the contribution of enriched entry descriptions. Median query-processing time is 13.37 seconds with six concurrent queries.
在包含 500 个动作描述查询的词典基准测试中，该系统的 Hit@1 达到 94.0%，Hit@9 达到 97.4%，平均倒数排名（MRR）为 0.9540。重排序将 Hit@1 从 71.8% 提升至 94.0%，组件分析也验证了扩充词条描述的贡献。在六个并发查询的情况下，中位查询处理时间为 13.37 秒。
By connecting everyday movement descriptions to documented signs and meanings, SignTrace provides a practical tool for identifying unfamiliar signs. Dictionary-derived wording and prior selection within the benchmark limit generalization to descriptions independently produced by users.
通过将日常动作描述与已记录的手势和含义相关联，SignTrace 为识别陌生手势提供了一种实用工具。由于基准测试中的措辞源自词典且经过预先筛选，其对用户独立生成的描述的泛化能力仍有一定局限。