Human Intention Recognition for Safe Robot Action Planning Using Head Pose

Orsag, Luka; Stipančić, Tomislav; Koren, Leon; Posavec, Karlo

doi:10.1007/978-3-031-17618-0_23

Cited by 1 publication

(3 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…False positives themselves do not represent a critical problem; rather, their intermittent behavior during short time intervals are problematic, since they can be regarded as true positive detections. In the future, these problems will be mitigated by post-processing and filtering, together with a multimodal information fusion, based on the recognition of human worker intention, as shown in [ 29 ].…”

Section: Discussionmentioning

confidence: 99%

“…The work presented in this paper, together with the work discussed in [ 29 ], represents the building blocks for efficient HRC, the final goal of this research, in which the robot and the worker can closely cooperate.…”

Section: Discussionmentioning

confidence: 99%

“…For example, the movements of the worker’s body can be associated with a possible collision with the robot [ 28 ]. In [ 29 ], the authors proposed an addition to a safety framework using a worker intention recognition method, based on head pose information. In [ 30 ], the authors described a behavior recognition system, based on a real-time stereo face-tracking and gaze detection designed to measure the head pose and gaze direction simultaneously.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Towards a Safe Human–Robot Collaboration Using Information on Human Worker Activity

Orsag

Stipančić

Koren

2023

Sensors

View full text Add to dashboard Cite

Most industrial workplaces involving robots and other apparatus operate behind the fences to remove defects, hazards, or casualties. Recent advancements in machine learning can enable robots to co-operate with human co-workers while retaining safety, flexibility, and robustness. This article focuses on the computation model, which provides a collaborative environment through intuitive and adaptive human–robot interaction (HRI). In essence, one layer of the model can be expressed as a set of useful information utilized by an intelligent agent. Within this construction, a vision-sensing modality can be broken down into multiple layers. The authors propose a human-skeleton-based trainable model for the recognition of spatiotemporal human worker activity using LSTM networks, which can achieve a training accuracy of 91.365%, based on the InHARD dataset. Together with the training results, results related to aspects of the simulation environment and future improvements of the system are discussed. By combining human worker upper body positions with actions, the perceptual potential of the system is increased, and human–robot collaboration becomes context-aware. Based on the acquired information, the intelligent agent gains the ability to adapt its behavior according to its dynamic and stochastic surroundings.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%