Direction: Sensing, Recognition, and Security

Everyday Behavior and Activity Recognition

Stats are based on currently indexed question data; missing sources remain visible.

items

sources

2024

latest

All questions

48 items

All Research questions Practical problems

helpResearch questionEveryday Behavior and Activity Recognition

How can non-visual egocentric smartphone sensors enable efficient logging of activities of daily living (ADL)?

UbiComp '24Multimodal Daily Life Logging Using Non-Visual Egocentric Sensors on Smartphones

helpResearch questionEveryday Behavior and Activity Recognition

How can multimodal fusion and self-supervised learning improve performance and generalization of non-visual ADL recognition?

UbiComp '24Multimodal Daily Life Logging Using Non-Visual Egocentric Sensors on Smartphones

helpResearch questionEveryday Behavior and Activity Recognition

Can non-visual ADL recognition replace camera-based methods to reduce privacy risks and energy consumption?

UbiComp '24Multimodal Daily Life Logging Using Non-Visual Egocentric Sensors on Smartphones

lightbulbPractical problemEveryday Behavior and Activity Recognition

Users worry about privacy and security with camera sensors, and existing methods struggle to capture complex behaviors.

UbiComp '24Multimodal Daily Life Logging Using Non-Visual Egocentric Sensors on Smartphones

helpResearch questionEveryday Behavior and Activity Recognition

In which practical special education scenarios can AI-assisted behavior recognition systems support autistic children?

UbiComp '24Exploring Potential Application Areas of Artificial Intelligence-Infused System for Engagement Recognition: Insights from Special Education Experts

helpResearch questionEveryday Behavior and Activity Recognition

Can multimodal data (e.g., visual and physiological signals) accurately identify engagement states of autistic children?

UbiComp '24Exploring Potential Application Areas of Artificial Intelligence-Infused System for Engagement Recognition: Insights from Special Education Experts

helpResearch questionEveryday Behavior and Activity Recognition

How can AI systems combining expert experience optimize teaching interventions in special education?

UbiComp '24Exploring Potential Application Areas of Artificial Intelligence-Infused System for Engagement Recognition: Insights from Special Education Experts

lightbulbPractical problemEveryday Behavior and Activity Recognition

Special education teachers struggle to assess autistic children's engagement in real time and optimize interventions.

UbiComp '24Exploring Potential Application Areas of Artificial Intelligence-Infused System for Engagement Recognition: Insights from Special Education Experts

helpResearch questionEveryday Behavior and Activity Recognition

How can ChatGPT guide inexperienced researchers through sensor-based human activity recognition tasks?

UbiComp '24How ChatGPT Assists Novices in Human Activity Recognition

helpResearch questionEveryday Behavior and Activity Recognition

What specific assistance can ChatGPT provide beginners in handling missing sensor data and class imbalance?

UbiComp '24How ChatGPT Assists Novices in Human Activity Recognition

helpResearch questionEveryday Behavior and Activity Recognition

How can ChatGPT optimize sensor data fusion and model selection when beginners use classical machine learning models?

UbiComp '24How ChatGPT Assists Novices in Human Activity Recognition

lightbulbPractical problemEveryday Behavior and Activity Recognition

Beginners find it difficult to develop human activity recognition algorithms using machine learning.

UbiComp '24How ChatGPT Assists Novices in Human Activity Recognition

helpResearch questionEveryday Behavior and Activity Recognition

How can IMUs mounted on shin guards detect specific actions of soccer players?

UbiComp '24A Method for Detecting Soccer Player Movements Using Inertial Measurement Units Attached to Shin Guards

helpResearch questionEveryday Behavior and Activity Recognition

How does IMU-based action recognition compare to traditional camera systems in accuracy and cost?

UbiComp '24A Method for Detecting Soccer Player Movements Using Inertial Measurement Units Attached to Shin Guards

helpResearch questionEveryday Behavior and Activity Recognition

What improvements can enhance the generalizability and accuracy of IMU-based action recognition algorithms?

UbiComp '24A Method for Detecting Soccer Player Movements Using Inertial Measurement Units Attached to Shin Guards

lightbulbPractical problemEveryday Behavior and Activity Recognition

Amateur soccer players lack low-cost tools for motion analysis.

UbiComp '24A Method for Detecting Soccer Player Movements Using Inertial Measurement Units Attached to Shin Guards

helpResearch questionEveryday Behavior and Activity Recognition

How can multiple wearable sensor data streams be used for daily activity tracking and natural language description generation?

UbiComp '24Sensor2Text: Enabling Natural Language Interactions for Daily Activity Tracking Using Wearable Sensors

helpResearch questionEveryday Behavior and Activity Recognition

How can cross-modal learning from sensor data improve information density and activity recognition accuracy?

UbiComp '24Sensor2Text: Enabling Natural Language Interactions for Daily Activity Tracking Using Wearable Sensors

helpResearch questionEveryday Behavior and Activity Recognition

Can natural language interaction with users be achieved while preserving privacy?

UbiComp '24Sensor2Text: Enabling Natural Language Interactions for Daily Activity Tracking Using Wearable Sensors

lightbulbPractical problemEveryday Behavior and Activity Recognition

Video-based activity monitoring exposes privacy and performs poorly in low-light conditions.

UbiComp '24Sensor2Text: Enabling Natural Language Interactions for Daily Activity Tracking Using Wearable Sensors

helpResearch questionEveryday Behavior and Activity Recognition

How can a fitness action classification model based on a single millimeter-wave radar overcome viewpoint dependency?

UbiComp '24View-agnostic Human Exercise Cataloging with Single MmWave Radar

helpResearch questionEveryday Behavior and Activity Recognition

How can self-supervised learning improve action classification accuracy under low-viewpoint conditions?

UbiComp '24View-agnostic Human Exercise Cataloging with Single MmWave Radar

helpResearch questionEveryday Behavior and Activity Recognition

How can millimeter-wave radar signals improve the real-time accuracy of fitness action repetition counting?

UbiComp '24View-agnostic Human Exercise Cataloging with Single MmWave Radar

lightbulbPractical problemEveryday Behavior and Activity Recognition

Users struggle to accurately record exercise type and repetition counts with privacy-friendly technology during workouts.

UbiComp '24View-agnostic Human Exercise Cataloging with Single MmWave Radar

helpResearch questionEveryday Behavior and Activity Recognition

How can neuromorphic cameras and event tags enable long-range, battery-free wireless sensing?

UbiComp '24NeuroCamTags: Long-Range, Battery-free, Wireless Sensing with Neuromorphic Cameras

helpResearch questionEveryday Behavior and Activity Recognition

How can low-power interactive smart environment sensing systems be designed based on energy harvesting and visible light communication?

UbiComp '24NeuroCamTags: Long-Range, Battery-free, Wireless Sensing with Neuromorphic Cameras

helpResearch questionEveryday Behavior and Activity Recognition

Can neuromorphic cameras provide higher accuracy and privacy protection than traditional sensors in activity recognition?

UbiComp '24NeuroCamTags: Long-Range, Battery-free, Wireless Sensing with Neuromorphic Cameras

lightbulbPractical problemEveryday Behavior and Activity Recognition

Sensors in smart environments have high maintenance costs and insufficient privacy protection.

UbiComp '24NeuroCamTags: Long-Range, Battery-free, Wireless Sensing with Neuromorphic Cameras

helpResearch questionEveryday Behavior and Activity Recognition

How can wearable sensors efficiently interpret multidimensional human motion signals?

UbiComp '23ViSig: Automatic Interpretation of Visual Body Signals Using On-Body Sensors

helpResearch questionEveryday Behavior and Activity Recognition

Can multi-sensor fusion address limitations of existing methods in recognizing complex movements?

UbiComp '23ViSig: Automatic Interpretation of Visual Body Signals Using On-Body Sensors

helpResearch questionEveryday Behavior and Activity Recognition

How can a general software framework be designed to meet human action recognition needs across diverse real-world scenarios?

UbiComp '23ViSig: Automatic Interpretation of Visual Body Signals Using On-Body Sensors

lightbulbPractical problemEveryday Behavior and Activity Recognition

Vision technologies are sensitive to lighting and occlusion, making precise human signal recognition difficult in industrial safety and motion scoring.

UbiComp '23ViSig: Automatic Interpretation of Visual Body Signals Using On-Body Sensors

helpResearch questionEveryday Behavior and Activity Recognition

How can large language models (e.g., ChatGPT) advise sensor placement in human activity recognition?

UbiComp '23Toward Pioneering Sensors and Features Using Large Language Models in Human Activity Recognition

helpResearch questionEveryday Behavior and Activity Recognition

Can large language models improve feature computation methods in human activity recognition?

UbiComp '23Toward Pioneering Sensors and Features Using Large Language Models in Human Activity Recognition

helpResearch questionEveryday Behavior and Activity Recognition

Can sensor and feature combinations suggested by large language models improve predictive model performance and reduce resource consumption?

UbiComp '23Toward Pioneering Sensors and Features Using Large Language Models in Human Activity Recognition

lightbulbPractical problemEveryday Behavior and Activity Recognition

Engineers spend significant time on manual sensor placement and feature design and depend on specialized expertise.

UbiComp '23Toward Pioneering Sensors and Features Using Large Language Models in Human Activity Recognition

helpResearch questionEveryday Behavior and Activity Recognition

How can sparse-layout tactile sensors enable room-scale human activity sensing?

UIST '22Flexel: A Modular Floor Interface for Room-Scale Tactile Sensing

helpResearch questionEveryday Behavior and Activity Recognition

How can modular floor interfaces enable footprint tracking, foot gesture recognition, and object localization at low cost?

UIST '22Flexel: A Modular Floor Interface for Room-Scale Tactile Sensing

helpResearch questionEveryday Behavior and Activity Recognition

How can floor sensor density be optimized to reduce cost while maintaining performance?

UIST '22Flexel: A Modular Floor Interface for Room-Scale Tactile Sensing

lightbulbPractical problemEveryday Behavior and Activity Recognition

Current floor systems cannot intelligently sense human activity and inadequately protect privacy.

UIST '22Flexel: A Modular Floor Interface for Room-Scale Tactile Sensing

helpResearch questionEveryday Behavior and Activity Recognition

How do fake data injection attacks in human-centered sensing applications (e.g., health monitoring, activity recognition) affect the integrity of sensor data?

UbiComp '22Countermeasures Against False Data Injection Attacks in Human-Centric Sensing Applications

helpResearch questionEveryday Behavior and Activity Recognition

How can multi-instance learning (MIL) and Siamese neural networks be used to detect these attacks?

UbiComp '22Countermeasures Against False Data Injection Attacks in Human-Centric Sensing Applications

helpResearch questionEveryday Behavior and Activity Recognition

How can an efficient, low-overhead fake data injection attack detection framework be designed for resource-constrained smart devices?

UbiComp '22Countermeasures Against False Data Injection Attacks in Human-Centric Sensing Applications

lightbulbPractical problemEveryday Behavior and Activity Recognition

Wearable device users may suffer incorrect health diagnoses or privacy threats due to fake data injection.

UbiComp '22Countermeasures Against False Data Injection Attacks in Human-Centric Sensing Applications

helpResearch questionEveryday Behavior and Activity Recognition

How can LFI sensors and IMUs be combined in smart glasses to achieve accurate recognition of complex human activities?

UbiComp '21A CNN-based Human Activity Recognition System Combining a Laser Feedback Interferometry Eye Movement Sensor and an IMU for Context-aware Smart Glasses

helpResearch questionEveryday Behavior and Activity Recognition

How can 1D-CNN deep learning methods optimize classification performance of mixed activities (e.g., cognitive and physical activities)?

UbiComp '21A CNN-based Human Activity Recognition System Combining a Laser Feedback Interferometry Eye Movement Sensor and an IMU for Context-aware Smart Glasses

helpResearch questionEveryday Behavior and Activity Recognition

Compared to existing single-sensing activity recognition methods, in which scenarios can the new approach improve recognition efficiency?

UbiComp '21A CNN-based Human Activity Recognition System Combining a Laser Feedback Interferometry Eye Movement Sensor and an IMU for Context-aware Smart Glasses

lightbulbPractical problemEveryday Behavior and Activity Recognition

Touch and voice interaction limit smart glasses immersion and make complex activity recognition inconvenient.

UbiComp '21A CNN-based Human Activity Recognition System Combining a Laser Feedback Interferometry Eye Movement Sensor and an IMU for Context-aware Smart Glasses

Adjacent categories

Sensing, Recognition, and Security

Everyday Behavior and Activity Recognition

All questions

Related papers

NeuroCamTags: Long-Range, Battery-free, Wireless Sensing with Neuromorphic Cameras

ViSig: Automatic Interpretation of Visual Body Signals Using On-Body Sensors

Flexel: A Modular Floor Interface for Room-Scale Tactile Sensing

Adjacent categories

Gesture and Pose Sensing Model Performance and Accuracy

Wearable Health Devices and Everyday Sensing

Speech, Face, and Body Pose Input

Biometric Authentication and Secure Identification

HAR Model Robustness, Generalization, and Data Augmentation

Wearable Health, Activity, and Behavior Monitoring

IMU Inertial Sensing

Mobile Touch and Micro-Gesture Input