arXiv cs.CL· Qingming Li, Ruiyan Sun, Satoshi Nakamura·· 12 小时前AI 评分24
NarraSpace：面向看图描述痴呆症分类的多模态场景关联证据
NarraSpace: Multimodal Scene-Grounded Evidence for Picture-Description Dementia Classification
AI 导读
NarraSpace 将描述文本中的词语和语音时间对齐到图片命名元素，并按预设场景事件检索话语，用冻结语言模型对比阿尔茨海默病与健康对照样本下的事件文本。在 ADReSS 官方 48 人测试集上，固定支持向量机五十个随机种子平均准确率达 80.88%±2.56%，平均概率给出 81.25% 准确率和 91.32% AUC。
正文 · 原文
Abstract:Picture-description dementia classification uses a shared visual scene, but a transcript-level prediction alone does not identify the scene evidence underlying its inputs. NarraSpace links words and speech times to named picture elements and retrieves utterances for predefined scene events. Narrative-spatial descriptors summarize coverage and temporal organization. A frozen language model contrasts each event's text under examples from Alzheimer's disease and healthy-control training participants, alongside transcript surprisal and coverage measurements. Retained source records link event scores to their source utterances, speech intervals, and lexical matches. On the official 48-participant test set of the Alzheimer's Dementia Recognition through Spontaneous Speech benchmark, a fixed support vector classifier achieves $80.88\%\pm2.56\%$ mean accuracy over fifty seeds; averaged probabilities yield $81.25\%$ accuracy and $91.32\%$ area under the receiver operating characteristic curve. NarraSpace makes the construction of scene-linked classification inputs explicit and inspectable by connecting event-indexed language-model measurements to their source utterances, speech timing, and extraction records. Such source-linked measurements could support clinician review of picture descriptions by presenting quantitative summaries alongside the original speech evidence.
| Comments: | 30 pages, 3 figures |
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2610.28616 [cs.CL] |
| (or arXiv:2610.28616v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2610.28616 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Qingming Li [view email]
[v1]
Tue, 20 Oct 2026 17:55:44 UTC (988 KB)
来源：arXiv cs.CL · arxiv.org