Context-aware Human Intent Inference for Improving Human Machine Cooperation

TL;DR

This paper proposes a computational framework that uses heterogeneous body and brain sensors to accurately infer human intentions in real-time, aiming to enhance human-machine cooperation and activity prediction.

Contribution

It introduces a novel approach combining cognitive and physiological sensor data for real-time human intent recognition, advancing human-machine interaction capabilities.

Findings

Framework enables reliable real-time intent inference

Improves accuracy of human activity prediction

Enhances human-machine cooperation

Abstract

The ability of human beings to precisely recog- nize others intents is a significant mental activity in reasoning about actions, such as, what other people are doing and what they will do next. Recent research has revealed that human intents could be inferred by measuring human cognitive activities through heterogeneous body and brain sensors (e.g., sensors for detecting physiological signals like ECG, brain signals like EEG and IMU sensors like accelerometers and gyros etc.). In this proposal, we aim at developing a computa- tional framework for enabling reliable and precise real-time human intent recognition by measuring human cognitive and physiological activities through the heterogeneous body and brain sensors for improving human machine interactions, and serving intent-based human activity prediction.