缩小零样本HAR中的模态差距：基于IMU数据的对比训练与可分性优化原型
Closing the Modality Gap in Zero-Shot HAR: Contrastive Training and Separability-Optimized Prototypes on IMU Data
AI总结：
针对IMU基零样本人体活动识别中的模态差距问题，提出对比训练与描述性原型结合的方法，在PAMAP2数据集上实现73.2%准确率和0.583宏F1，并指出宏F1更适合作为评估指标。
摘要：
基于惯性测量单元（IMU）的人体活动识别（HAR）中的零样本学习（ZSL）面临一个核心挑战：弥合传感器嵌入与语义类表示之间的差距。我们在PAMAP2数据集上系统评估了三种推理方法与两种训练流程组合的七种配置，使用14个已知和4个未知活动类别，并保留受试者108和109用于测试。我们发现模态差距是一个由编码器目标决定的训练时现象。使用标签名称的Sentence-BERT原型进行交叉熵训练的时间卷积网络（TCN）产生的传感器嵌入与对应文本原型的平均余弦相似度为0.30，而将标签名称原型目标替换为判别性活动描述后，该值提升至0.69。这种对齐改进在所有三种推理方法中一致迁移。最强的结果结合了对比训练与反向softmax校正，在未知类别上达到73.2%的准确率和0.583的宏F1，而标签名称基线仅为58.3%准确率和0.34宏F1。另一个发现是，更丰富的文本描述降低了Sentence-BERT空间中原型间的可分性，因为共享的生物力学词汇导致语言模型压缩了原型云。只要原型描述保留足够的判别性词汇，这种效应不会抵消对比对齐的好处。我们还证明，当测试集类别分布不平衡时，总体准确率是一个误导性的主要指标，并推荐宏平均F1作为ZSL-HAR基准的标准报告指标。
英文摘要：
Zero-shot learning (ZSL) for inertial measurement unit (IMU)-based human activity recognition (HAR) faces a central challenge: bridging the gap between sensor embeddings and semantic class representations. We systematically evaluate seven configurations combining three inference methods with two training pipelines on the PAMAP2 dataset, using 14 seen and 4 unseen activity classes with subjects 108 and 109 held out for testing. We find that the modality gap is a training-time phenomenon governed by the encoder objective. A temporal convolutional network (TCN) trained with cross-entropy over label-name Sentence- BERT prototypes yields sensor embeddings with a mean cosine similarity of 0.30 to the corresponding text prototypes, while replacing the label-name prototype targets with discriminative activity descriptions raises this to 0.69. This alignment improvement transfers consistently across all three inference methods. The strongest result combines contrastive training with inverted softmax correction, achieving 73.2% accuracy and 0.583 macro F1 on unseen classes, compared to 58.3% accuracy and 0.34 macro F1 for the label-name baseline. A secondary finding is that richer text descriptions reduce inter-prototype separability in Sentence-BERT space, because shared biomechanical vocabulary causes the language model to compress the prototype cloud. This effect does not negate the benefits of contrastive alignment provided prototype descriptions retain sufficient discriminative vocabulary. We also demonstrate that overall accuracy is a misleading primary metric when test-set class distributions are imbalanced, and recommend macro-averaged F1 as the standard reporting metric for ZSL-HAR benchmarks.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
