help研究问题HAR模型鲁棒性、泛化与数据增强Kolmogorov-Arnold网络(KANs)能否更高效地对IMU传感器数据进行特征提取以支持人类活动识别?UbiComp '24Initial Investigation of Kolmogorov-Arnold Networks (KANs) as Feature Extractors for IMU Based Human Activity Recognition
help研究问题HAR模型鲁棒性、泛化与数据增强单通道、多通道与跨通道特征对KAN架构在人类活动识别中的表现有何影响?UbiComp '24Initial Investigation of Kolmogorov-Arnold Networks (KANs) as Feature Extractors for IMU Based Human Activity Recognition
help研究问题HAR模型鲁棒性、泛化与数据增强在相比CNN的特征提取中,KAN在参数效率和分类准确性上是否具有优势?UbiComp '24Initial Investigation of Kolmogorov-Arnold Networks (KANs) as Feature Extractors for IMU Based Human Activity Recognition
lightbulb现实痛点HAR模型鲁棒性、泛化与数据增强现有方法对IMU数据特征提取效率低,导致活动识别性能受限。UbiComp '24Initial Investigation of Kolmogorov-Arnold Networks (KANs) as Feature Extractors for IMU Based Human Activity Recognition
help研究问题HAR模型鲁棒性、泛化与数据增强主动声学技术能否通过智能眼镜精确监测并识别多样的日常活动?UbiComp '24ActSonic: Recognizing Everyday Activities from Inaudible Acoustic Wave Around the Body
help研究问题HAR模型鲁棒性、泛化与数据增强如何利用深度学习和自监督学习,从无标签数据中提取有效特征以提高活动识别性能?UbiComp '24ActSonic: Recognizing Everyday Activities from Inaudible Acoustic Wave Around the Body
help研究问题HAR模型鲁棒性、泛化与数据增强主动声学传感能否在能耗低、保护隐私的情况下,胜任现有摄像头或惯性传感器无法处理的活动识别任务?UbiComp '24ActSonic: Recognizing Everyday Activities from Inaudible Acoustic Wave Around the Body
help研究问题HAR模型鲁棒性、泛化与数据增强如何在未知目标数据集中实现传感器基础的人类活动识别的域泛化能力?UbiComp '24CrossHAR: Generalizing Cross-dataset Human Activity Recognition via Hierarchical Self-Supervised Pretraining
help研究问题HAR模型鲁棒性、泛化与数据增强物理知识驱动的数据增强和分层自监督学习如何提高IMU(惯性测量单元)数据模型的泛化性能?UbiComp '24CrossHAR: Generalizing Cross-dataset Human Activity Recognition via Hierarchical Self-Supervised Pretraining
help研究问题HAR模型鲁棒性、泛化与数据增强如何最好地利用IMU未标注数据以解决域转移问题?UbiComp '24CrossHAR: Generalizing Cross-dataset Human Activity Recognition via Hierarchical Self-Supervised Pretraining
lightbulb现实痛点HAR模型鲁棒性、泛化与数据增强智能设备中人类活动识别模型在新场景下经常失效。UbiComp '24CrossHAR: Generalizing Cross-dataset Human Activity Recognition via Hierarchical Self-Supervised Pretraining
help研究问题HAR模型鲁棒性、泛化与数据增强身体中心的电场感应如何提高人体活动识别的准确性和可靠性?UbiComp '24Body-Centric Capacitance or Electric Field Sensing for Human Activity Recognition and Human-Computer Interaction: A Comprehensive Review
help研究问题HAR模型鲁棒性、泛化与数据增强体内/跨体电场感应的硬件设计和数据处理方法有哪些优势和局限性?UbiComp '24Body-Centric Capacitance or Electric Field Sensing for Human Activity Recognition and Human-Computer Interaction: A Comprehensive Review
help研究问题HAR模型鲁棒性、泛化与数据增强如何通过电场感应与其他传感器的融合来改善信号抗干扰性和算法适用性?UbiComp '24Body-Centric Capacitance or Electric Field Sensing for Human Activity Recognition and Human-Computer Interaction: A Comprehensive Review
lightbulb现实痛点HAR模型鲁棒性、泛化与数据增强现有的人体活动识别技术易受环境影响且能耗较高。UbiComp '24Body-Centric Capacitance or Electric Field Sensing for Human Activity Recognition and Human-Computer Interaction: A Comprehensive Review
help研究问题HAR模型鲁棒性、泛化与数据增强如何构建大规模、高质量的人类活动识别(HAR)数据集来支持复杂行为的识别?UbiComp '2412th International Workshop on Human Activity Sensing Corpus and Applications (HASCA)
help研究问题HAR模型鲁棒性、泛化与数据增强哪些新的开放方法(如终身学习)可以增强活动识别方法的鲁棒性和适应性?UbiComp '2412th International Workshop on Human Activity Sensing Corpus and Applications (HASCA)
help研究问题HAR模型鲁棒性、泛化与数据增强如何通过国际协作和社区驱动的努力推进活动识别数据集的开发和共享?UbiComp '2412th International Workshop on Human Activity Sensing Corpus and Applications (HASCA)
lightbulb现实痛点HAR模型鲁棒性、泛化与数据增强现有算法难以在真实复杂环境中准确识别人类活动。UbiComp '2412th International Workshop on Human Activity Sensing Corpus and Applications (HASCA)
help研究问题HAR模型鲁棒性、泛化与数据增强在传感器数据随机丢失的情况下,如何增强基于多传感器数据的人类活动识别的准确性和鲁棒性?UbiComp '24Sussex-Huawei Locomotion Recognition Using Machine Learning and Deep Learning with Multi-sensor Data
help研究问题HAR模型鲁棒性、泛化与数据增强通过什么方法可以有效处理随机丢失的多模态传感器数据,实现高精度的人类活动分类?UbiComp '24Sussex-Huawei Locomotion Recognition Using Machine Learning and Deep Learning with Multi-sensor Data
help研究问题HAR模型鲁棒性、泛化与数据增强三分支卷积神经网络(CNN)如何在输送缺失数据的情况下保持分类性能?UbiComp '24Sussex-Huawei Locomotion Recognition Using Machine Learning and Deep Learning with Multi-sensor Data
lightbulb现实痛点HAR模型鲁棒性、泛化与数据增强健康监测等场景中传感器数据丢失导致活动识别效果下降。UbiComp '24Sussex-Huawei Locomotion Recognition Using Machine Learning and Deep Learning with Multi-sensor Data
help研究问题HAR模型鲁棒性、泛化与数据增强非专家用户如何有效定义和调试基于知识的人类活动识别(HAR)系统?UbiComp '24exHAR: An Interface for Helping Non-experts Develop and Debug Knowledge-based Human Activity Recognition Systems
help研究问题HAR模型鲁棒性、泛化与数据增强在支持非专家用户调试时,'为什么'和'如果'类型的解释机制能带来多大帮助?UbiComp '24exHAR: An Interface for Helping Non-experts Develop and Debug Knowledge-based Human Activity Recognition Systems
help研究问题HAR模型鲁棒性、泛化与数据增强用户对基于知识的HAR系统中规则定义和错误修正的认知过程是怎样的?UbiComp '24exHAR: An Interface for Helping Non-experts Develop and Debug Knowledge-based Human Activity Recognition Systems
lightbulb现实痛点HAR模型鲁棒性、泛化与数据增强非专家用户难以灵活定义和调试个性化活动识别规则。UbiComp '24exHAR: An Interface for Helping Non-experts Develop and Debug Knowledge-based Human Activity Recognition Systems
help研究问题HAR模型鲁棒性、泛化与数据增强在传感器数据缺失的情况下,人类活动识别的准确性如何提升?UbiComp '24Ensemble Learning Approach for Human Activity Recognition Involving Missing Sensor Data
help研究问题HAR模型鲁棒性、泛化与数据增强深度学习与传统机器学习(如XGBoost和LightGBM)结合后,能否改善不同缺失传感器场景下的分类性能?UbiComp '24Ensemble Learning Approach for Human Activity Recognition Involving Missing Sensor Data
help研究问题HAR模型鲁棒性、泛化与数据增强哪种多模型融合策略最适合处理多源数据中传感器缺失问题?UbiComp '24Ensemble Learning Approach for Human Activity Recognition Involving Missing Sensor Data
lightbulb现实痛点HAR模型鲁棒性、泛化与数据增强用户在传感器数据缺失时无法准确识别活动,影响设备功能。UbiComp '24Ensemble Learning Approach for Human Activity Recognition Involving Missing Sensor Data
help研究问题HAR模型鲁棒性、泛化与数据增强扩散模型能否有效应用于一维传感器信号的人类活动识别(HAR)任务?UbiComp '24Analysis of Human Activity Recognition by Diffusion Models
help研究问题HAR模型鲁棒性、泛化与数据增强如何设计基于扩散模型的架构以优化一维信号的分类性能?UbiComp '24Analysis of Human Activity Recognition by Diffusion Models
help研究问题HAR模型鲁棒性、泛化与数据增强比较于现有方法(如Transformer或GAN等),基于扩散模型的分类器在哪些方面表现优异?UbiComp '24Analysis of Human Activity Recognition by Diffusion Models
lightbulb现实痛点HAR模型鲁棒性、泛化与数据增强用户佩戴的多传感设备生成的大量传感信号难以被精准高效识别。UbiComp '24Analysis of Human Activity Recognition by Diffusion Models
help研究问题HAR模型鲁棒性、泛化与数据增强如何将时间敏感的适配器(Temporal-Informative Adapter)从视觉数据领域迁移到惯性数据领域,并提升特征提取能力?UbiComp '24TA-DA! - Improving Activity Recognition Using Temporal Adapters and Data Augmentation
help研究问题HAR模型鲁棒性、泛化与数据增强数据增强方法(如轴混淆、归一化、噪声添加)如何改善活动识别模型的泛化性能?UbiComp '24TA-DA! - Improving Activity Recognition Using Temporal Adapters and Data Augmentation
help研究问题HAR模型鲁棒性、泛化与数据增强在惯性数据活动识别中,时间敏感的适配器与数据增强相结合是否能获得更好的模型性能?UbiComp '24TA-DA! - Improving Activity Recognition Using Temporal Adapters and Data Augmentation
lightbulb现实痛点HAR模型鲁棒性、泛化与数据增强用户使用智能设备时,活动识别准确性不足,影响健康监测和运动追踪。UbiComp '24TA-DA! - Improving Activity Recognition Using Temporal Adapters and Data Augmentation
help研究问题HAR模型鲁棒性、泛化与数据增强如何通过自监督学习从可穿戴传感器的数据中提取有效特征?UbiComp '24Solving the Sensor-Based Activity Recognition Problem (SOAR): Self-Supervised, Multi-Modal Recognition of Activities from Wearable Sensors
help研究问题HAR模型鲁棒性、泛化与数据增强多模态数据(如视频、音频与传感器数据)的融合能否提高人类活动识别的准确性?UbiComp '24Solving the Sensor-Based Activity Recognition Problem (SOAR): Self-Supervised, Multi-Modal Recognition of Activities from Wearable Sensors
help研究问题HAR模型鲁棒性、泛化与数据增强在缺乏大规模标注数据的情况下,如何提升传感器领域模型的适用性?UbiComp '24Solving the Sensor-Based Activity Recognition Problem (SOAR): Self-Supervised, Multi-Modal Recognition of Activities from Wearable Sensors
help研究问题HAR模型鲁棒性、泛化与数据增强在户外运动中,多模态数据(惯性传感器和自我视角视频)如何结合以改进人类活动识别?UbiComp '24WEAR: An Outdoor Sports Dataset for Wearable and Egocentric Activity Recognition
help研究问题HAR模型鲁棒性、泛化与数据增强现有的单模态时间动作定位模型能否扩展应用于惯性数据分析?UbiComp '24WEAR: An Outdoor Sports Dataset for Wearable and Egocentric Activity Recognition
help研究问题HAR模型鲁棒性、泛化与数据增强多传感器融合的数据集设计如何支持动态的户外运动场景?UbiComp '24WEAR: An Outdoor Sports Dataset for Wearable and Egocentric Activity Recognition
lightbulb现实痛点HAR模型鲁棒性、泛化与数据增强现有活动识别数据集缺乏户外场景的多模态支持。UbiComp '24WEAR: An Outdoor Sports Dataset for Wearable and Egocentric Activity Recognition
help研究问题HAR模型鲁棒性、泛化与数据增强如何提高传感器中基于深度学习的人体活动识别模型的跨主体通用性?UbiComp '24AutoAugHAR: Automated Data Augmentation for Sensor-based Human Activity Recognition
help研究问题HAR模型鲁棒性、泛化与数据增强自动优化的两阶段数据增强方法如何有效平衡数据多样性和标签语义保持?UbiComp '24AutoAugHAR: Automated Data Augmentation for Sensor-based Human Activity Recognition
help研究问题HAR模型鲁棒性、泛化与数据增强在多模态传感器数据中,如何为不同模态设计适应性数据增强策略?UbiComp '24AutoAugHAR: Automated Data Augmentation for Sensor-based Human Activity Recognition
lightbulb现实痛点HAR模型鲁棒性、泛化与数据增强人体活动识别模型在处理不同用户的传感器数据时表现不稳定。UbiComp '24AutoAugHAR: Automated Data Augmentation for Sensor-based Human Activity Recognition