helpResearch question自监督学习与域泛化的活动识别方法
在智能手机多传感器数据缺失的情况下，如何有效识别用户的活动和交通方式？UbiComp '24Summary of SHL Challenge 2024: Motion Sensor-based Locomotion and Transportation Mode Recognition in Missing Data Scenarios
Direction: 传感、识别与安全
自监督学习与域泛化的活动识别方法
Stats are based on currently indexed question data; missing sources remain visible.
35
items
9
sources
2024
latest
All questions
35 items
helpResearch question自监督学习与域泛化的活动识别方法
缺失数据场景下，单分类模型和多模型（组合模型）策略的性能如何比较？UbiComp '24Summary of SHL Challenge 2024: Motion Sensor-based Locomotion and Transportation Mode Recognition in Missing Data Scenarios
helpResearch question自监督学习与域泛化的活动识别方法
传感器模态缺失的动态电量消耗场景中，哪些活动类别更易被准确识别？UbiComp '24Summary of SHL Challenge 2024: Motion Sensor-based Locomotion and Transportation Mode Recognition in Missing Data Scenarios
lightbulbPractical problem自监督学习与域泛化的活动识别方法
用户设备因节能模式或硬件限制，传感器数据丢失导致活动识别失准。UbiComp '24Summary of SHL Challenge 2024: Motion Sensor-based Locomotion and Transportation Mode Recognition in Missing Data Scenarios
helpResearch question自监督学习与域泛化的活动识别方法
如何解决多模态移动感知数据中训练数据与真实应用域之间的分布偏移问题？UbiComp '24
helpResearch question自监督学习与域泛化的活动识别方法
多分支对抗训练如何在无监督领域适配中有效应对多模态数据的复杂性？UbiComp '24
helpResearch question自监督学习与域泛化的活动识别方法
使用何种统计方法可以动态调整多模态数据的训练参数以提高适应性能？UbiComp '24
lightbulbPractical problem自监督学习与域泛化的活动识别方法
多模态传感技术跨用户和环境时，适应性差，标签稀缺。UbiComp '24
helpResearch question自监督学习与域泛化的活动识别方法
在数据模态缺失的情况下，如何可靠地检测交通方式？UbiComp '24Magnitude and Rotation Invariant Detection of Transportation Modes with Missing Data Modalities
helpResearch question自监督学习与域泛化的活动识别方法
如何通过特征提取实现传感器数据的旋转和幅值不变性？UbiComp '24Magnitude and Rotation Invariant Detection of Transportation Modes with Missing Data Modalities
helpResearch question自监督学习与域泛化的活动识别方法
传统机器学习方法如何应对传感器数据的分布变化问题？UbiComp '24Magnitude and Rotation Invariant Detection of Transportation Modes with Missing Data Modalities
lightbulbPractical problem自监督学习与域泛化的活动识别方法
移动设备识别用户交通方式时，受噪声和数据缺失影响大。UbiComp '24Magnitude and Rotation Invariant Detection of Transportation Modes with Missing Data Modalities
helpResearch question自监督学习与域泛化的活动识别方法
如何通过多模态传输生成多样化且高相关性的虚拟IMU数据以缓解人类活动识别中的数据稀缺问题？UbiComp '24IMUGPT 2.0: Cross-Modal Transfer Based on Natural Language for Sensor-Based Human Activity Recognition
helpResearch question自监督学习与域泛化的活动识别方法
如何使用自然语言模型和运动生成模型优化虚拟IMU数据的生成效率？UbiComp '24IMUGPT 2.0: Cross-Modal Transfer Based on Natural Language for Sensor-Based Human Activity Recognition
helpResearch question自监督学习与域泛化的活动识别方法
多样性测量和运动过滤如何提高生成数据的质量和实用性？UbiComp '24IMUGPT 2.0: Cross-Modal Transfer Based on Natural Language for Sensor-Based Human Activity Recognition
lightbulbPractical problem自监督学习与域泛化的活动识别方法
通过可穿戴设备收集活动数据成本高且隐私风险大，导致数据稀缺。UbiComp '24IMUGPT 2.0: Cross-Modal Transfer Based on Natural Language for Sensor-Based Human Activity Recognition
helpResearch question自监督学习与域泛化的活动识别方法
在跨域人类活动识别中，如何高效利用无标签数据以减少对标签的依赖？UbiComp '24Domain Adaptation in Human Activity Recognition through Self-Training
helpResearch question自监督学习与域泛化的活动识别方法
伪标签生成与多任务学习如何提升目标域的模型适应性？UbiComp '24Domain Adaptation in Human Activity Recognition through Self-Training
helpResearch question自监督学习与域泛化的活动识别方法
自我训练方法在跨环境数据集中的表现如何接近监督学习？UbiComp '24Domain Adaptation in Human Activity Recognition through Self-Training
lightbulbPractical problem自监督学习与域泛化的活动识别方法
智能手表等设备生成的大量无标签数据难以高效利用。UbiComp '24Domain Adaptation in Human Activity Recognition through Self-Training
helpResearch question自监督学习与域泛化的活动识别方法
能否通过改进伪标签生成方法，提高半监督学习在个性化HAR中的性能？UbiComp '23Personalized Federated Human Activity Recognition through Semi-supervised Learning and Enhanced Representation
helpResearch question自监督学习与域泛化的活动识别方法
如何通过增强表示学习来在数据异质性条件下提升HAR模型的泛化能力？UbiComp '23Personalized Federated Human Activity Recognition through Semi-supervised Learning and Enhanced Representation
lightbulbPractical problem自监督学习与域泛化的活动识别方法
用户的身体差异导致活动识别模型难以准确预测或满足个性化需求。UbiComp '23Personalized Federated Human Activity Recognition through Semi-supervised Learning and Enhanced Representation
helpResearch question自监督学习与域泛化的活动识别方法
如何通过层次聚类提升联邦学习在人类活动识别中的公平性和鲁棒性？UbiComp '23A Hierarchical Clustering-Based Personalized Federated Learning Framework for Robust and Fair Human Activity Recognition (FedCHAR)
helpResearch question自监督学习与域泛化的活动识别方法
在动态用户和数据环境中，如何实现联邦学习框架的可扩展性？UbiComp '23A Hierarchical Clustering-Based Personalized Federated Learning Framework for Robust and Fair Human Activity Recognition (FedCHAR)
helpResearch question自监督学习与域泛化的活动识别方法
如何利用模型更新方向的相似性降低数据异质性对联邦学习的影响并检测恶意节点？UbiComp '23A Hierarchical Clustering-Based Personalized Federated Learning Framework for Robust and Fair Human Activity Recognition (FedCHAR)
lightbulbPractical problem自监督学习与域泛化的活动识别方法
人类活动识别需保护隐私但受数据异质性和攻击影响性能降低。UbiComp '23A Hierarchical Clustering-Based Personalized Federated Learning Framework for Robust and Fair Human Activity Recognition (FedCHAR)
helpResearch question自监督学习与域泛化的活动识别方法
在活动识别中，如何在没有标签数据的情况下有效提取特征并实现分类？UbiComp '21Unsupervised Human Activity Representation Learning with Multi-task Deep Clustering
helpResearch question自监督学习与域泛化的活动识别方法
多任务深度聚类框架是否能通过联合优化实现更好的模型性能？UbiComp '21Unsupervised Human Activity Representation Learning with Multi-task Deep Clustering
helpResearch question自监督学习与域泛化的活动识别方法
利用伪标签生成的无监督方法在传感器数据上是否可以媲美监督学习？UbiComp '21Unsupervised Human Activity Representation Learning with Multi-task Deep Clustering
lightbulbPractical problem自监督学习与域泛化的活动识别方法
运动传感数据的标签标注成本高，隐私风险大。UbiComp '21Unsupervised Human Activity Representation Learning with Multi-task Deep Clustering
helpResearch question自监督学习与域泛化的活动识别方法
如何解决活动识别中因数据类别不平衡和训练-测试分布不一致导致的分类性能下降问题？UbiComp '21Triple-O for SHL Recognition Challenge: An Ensemble Framework for Multi-class Imbalance and Training-testing Distribution Inconsistency by OvO Binarization with Confidence Weight of One-class Classification
helpResearch question自监督学习与域泛化的活动识别方法
OvO（二分类）和单分类结合的框架如何提高多类分类任务的稳定性和精确性？UbiComp '21Triple-O for SHL Recognition Challenge: An Ensemble Framework for Multi-class Imbalance and Training-testing Distribution Inconsistency by OvO Binarization with Confidence Weight of One-class Classification
helpResearch question自监督学习与域泛化的活动识别方法
Triple-O框架在改善少数类预测性能方面与传统方法相比有哪些优势？UbiComp '21Triple-O for SHL Recognition Challenge: An Ensemble Framework for Multi-class Imbalance and Training-testing Distribution Inconsistency by OvO Binarization with Confidence Weight of One-class Classification
lightbulbPractical problem自监督学习与域泛化的活动识别方法
活动识别中常因数据类别不平衡和跨用户分布差异导致性能不佳。UbiComp '21Triple-O for SHL Recognition Challenge: An Ensemble Framework for Multi-class Imbalance and Training-testing Distribution Inconsistency by OvO Binarization with Confidence Weight of One-class Classification