Intelligent wearable systems: Opportunities and challenges in health and sports
Wearable devices, or wearables, designed to be attached to the human body, can gather
personalized real-time data and continuously monitor an individual's health status and …
personalized real-time data and continuously monitor an individual's health status and …
Personal llm agents: Insights and survey about the capability, efficiency and security
Since the advent of personal computing devices, intelligent personal assistants (IPAs) have
been one of the key technologies that researchers and engineers have focused on, aiming …
been one of the key technologies that researchers and engineers have focused on, aiming …
Mumu: Cooperative multitask learning-based guided multimodal fusion
Multimodal sensors (visual, non-visual, and wearable) can provide complementary
information to develop robust perception systems for recognizing activities accurately …
information to develop robust perception systems for recognizing activities accurately …
Patron: perspective-aware multitask model for referring expression grounding using embodied multimodal cues
Humans naturally use referring expressions with verbal utterances and nonverbal gestures
to refer to objects and events. As these referring expressions can be interpreted differently …
to refer to objects and events. As these referring expressions can be interpreted differently …
CAESAR: An embodied simulator for generating multimodal referring expression datasets
MM Islam, R Mirzaiee, A Gladstone… - Advances in Neural …, 2022 - proceedings.neurips.cc
Humans naturally use verbal utterances and nonverbal gestures to refer to various objects
(known as $\textit {referring expressions} $) in different interactional scenarios. As collecting …
(known as $\textit {referring expressions} $) in different interactional scenarios. As collecting …
IMPRINT: Interactional dynamics-aware motion prediction in teams using multimodal context
Robots are moving from working in isolation to working with humans as a part of human-
robot teams. In such situations, they are expected to work with multiple humans and need to …
robot teams. In such situations, they are expected to work with multiple humans and need to …
SMTDKD: A Semantic-Aware Multimodal Transformer Fusion Decoupled Knowledge Distillation Method for Action Recognition
Multimodal sensors, including vision sensors and wearable sensors, offer valuable
complementary information for accurate recognition tasks. Nonetheless, the heterogeneity …
complementary information for accurate recognition tasks. Nonetheless, the heterogeneity …
A State-of-the-Art Review of Computational Models for Analyzing Longitudinal Wearable Sensor Data in Healthcare
P Lago - arXiv preprint arXiv:2407.21665, 2024 - arxiv.org
Wearable devices are increasingly used as tools for biomedical research, as the continuous
stream of behavioral and physiological data they collect can provide insights about our …
stream of behavioral and physiological data they collect can provide insights about our …
ARCTIC: A knowledge distillation approach via attention-based relation matching and activation region constraint for RGB-to-Infrared videos action recognition
The recognition effect of existing infrared-based action recognition is greatly reduced when
clear appearance and texture are required. To address this limitation, the amalgamation of …
clear appearance and texture are required. To address this limitation, the amalgamation of …
MMBind: Unleashing the Potential of Distributed and Heterogeneous Data for Multimodal Learning in IoT
Multimodal sensing systems are increasingly prevalent in various real-world applications.
Most existing multimodal learning approaches heavily rely on training with a large amount of …
Most existing multimodal learning approaches heavily rely on training with a large amount of …