United States

A novel deep learning methodology for the assessment of ergonomic risks in any given work-related task with high accuracy and in real-time is presented. The proposed method takes as input a video sequence, extracts 3D skeletal features using VIBE algorithm and then processes these features to predict REBA risk scores for the entire body and individual body parts. Results on two well-known datasets and against other state-of-the-art methods verify the ability of the proposed method in predicting REBA ergonomic risk scores with high accuracy.

PETRA 2021

Towards Real-time Generalized Ergonomic Risk Assessment for the Prevention of Musculoskeletal Disorders

deep learning network

reba

ergonomics assessement

**About**
 The PErvasive Technologies Related to Assistive Environments (PETRA) conference is a highly interdisciplinary conference that focuses on computational and engineering approaches to improve the quality of life and enhance human performance in a wide range of settings, in the workplace, at home, in public spaces, urban environments, and other. Outcomes of this conference have a broad impact in application areas that include, manufacturing, transportation, healthcare, energy systems, security and safety, robotics, biomedicine, environment and conservation, and many others. PETRA brings together very different types of technologies to also address important social and healthcare issues for sensitive populations, such as the elderly, persons suffering from chronic conditions such as Alzheimer’s, Parkinson’s, Cerebral Palsy, and other disabilities or traumas.

 PETRA (which means “stone” in Greek) addresses important human living needs. From a technological perspective, PETRA motivates and inspires new scientific thrusts, innovations and discoveries to improve living environments, a particularly important issue, in an increasingly crowded and complex world where the need for inclusiveness and connectivity with the rest of the world is key. For the elderly in particular, PETRA addresses the fact that, as people grow older; they will increasingly rely on technology to be able to prolong staying in their homes and enjoy an independent life.

 The PETRA 2021 proceedings will be published in the ACM Digital Library as part of the ACM ICPS program. The ISBN number assigned to PETRA '21 is 978-1-4503-8792-7

The PErvasive Technologies Related to Assistive Environments (PETRA) conference is a highly interdisciplinary conference that focuses on computational and engineering approaches to improve the quality of life and enhance human performance in a wide range of settings, in the workplace, at home, in public spaces, urban environments, and other. Outcomes of this conference have a broad impact in application areas that include, manufacturing, transportation, healthcare, energy systems, security and safety, robotics, biomedicine, environment and conservation, and many others. PETRA brings together very different types of technologies to also address important social and healthcare issues for sensitive populations, such as the elderly, persons suffering from chronic conditions such as Alzheimer’s, Parkinson’s, Cerebral Palsy, and other disabilities or traumas.

workshop paper

Over the course of the coming decades, the societal goal of maximizing the number of older adults who experience positive trajectories of ageing has been established. To this end, innovative technologybased interventions tailored to older adults’ functional levels and focused on healthy lifestyles are considered imperative. The purpose of this work is to present a cluster of digital solutions aims at supporting and maintaining healthy and independent living for older adults, who face permanently or temporarily reduced functionality and capabilities.

Assistive Technologies for Supporting Wellbeing of Older Adults

Inferring the emotional state of an individual by viewing his/her facial expression seems to be present in all human cultures. Numerous
studies have shown that various changes in facial muscles determine the resulting facial expression. The analysis of images of faces
expressing emotional states promises to contribute to the quantification
of the claimed observations. Here, we use a suite of AI (artificial
intelligence) algorithms, along with ML (maximum likelihood) estimated distributions to quantify the shift in facial expression from
“neutral”→“fear” and “pain”→“pleasure”. The images are single
frames of five emotional states (neutral, fear, pain, pleasure, laugh)
expressed by actors and actresses in BDSM videos. We extract a feature vector for each image, dimension-reduce these feature vectors
by mapping them onto a two-dimensional manifold and calculate
the norms of the normalized displacement vectors for each emotional pair. We then find that the ML distributions of the norms are
Gamma-distributed and that the modes for each pair are different
for both males and females. We use Wilks lambda to determine
significance. We find that the distributions for the females are significantly different, but not for the males. The methodology we present
here has widespread applications: monitoring the emotional states
of humans in various settings; among these: determining whether
participants in BDSM and similar videos are indeed volunteering
their participation or are victims of criminal activity.

Visual Analysis of Emotions Using AI Image-Processing Software: Possible Male/Female Differences between the Emotion Pairs “Neutral”–“Fear” and “Pleasure”–“Pain”

As the field of robotics develops, so do the methods of interfacing with and controlling those robots. Many modern robots can be communicated with using commands that require neither speech nor touch. The two main motivations behind this trend are the assistance of people with disabilities and the desire for more natural Human-Robot Interaction (HRI). In allowing nonverbal communication with robots, accessibility of the systems increases to allow more people to interact with and benefit from robotics systems. Additionally, nonverbal communication provides more natural communication, which can lead to many benefits in HRI. This paper
provides an overview of existing technologies in touchless and nonverbal human-robot interfaces; it presents the most prevalent interface methods, their uses, and limitations. Discussions are provided on future directions for research.

Interfacing with Robots without the use of Touch or Speech

Force control is nowadays a mature technology and it is becoming more and more common in severed robotic applications. Despite this, an effective standardized and comprehensive method to assess force control performance does not exist. Defining the performance of a force-controlled system is not a trivial task, and one of the reason is that performance does not only depend on the dynamics of the robot, but also on the dynamics of the environment, which are usually uncertain and time-varying. Exerting a force on a soft environment is different from exerting a force on a rigid environment. Indeed, the same force-controlled robot can have different force responses in different environments. This paper proposes a methodology to standardize the benchmarking of force control algorithms applied to different actuation architectures and considering a wide set of possible environments and disturbances, defined by the application of interest.

Benchmarking Force Control Algorithms

We are presenting a conceptual design of a post-stroke rehabilitation
device, which is aimed for training the coupled movement
of wrist and fingers from flexion to extension in three modalities –
passive, active-assisted and active mode of operation. We will describe the configuration, actuation and control system of the
robotic device and provide the results of the mechanism’s mock-up
evaluation, therapists’ feedback on the design, future work and
development plans.

Robotic device for wrist and finger exercise

Automatic building extraction from satellite RGB images, is a low-cost alternative to perform important urban planning tasks. Yet, itis a challenging one, especially when natural and non-city block objects interfere in the semantic segmentation of algorithms that extract their key features. In this work we approach the automatic building extraction using a Convolution Neural Network based on the U-Net architecture. In contrast to existing approaches, it successfully encodes important features and decodes the buildings’ localization by requiring both reduced computational time and dataset size. We evaluate the U-Net’s performance using RGB images selected from the SpaceNet 1 dataset and the experimental results show an accuracy in building localization of 92.3%. Finally,favorable comparison with existing CNN approaches to hyper-spectral images targeting the SpaceNet 1 dataset, demonstrated its effectiveness.

Building Extraction from RGB Satellite Images using Deep Learning: A U-Net Approach

The presentation provides a novel method for reducing a multi-class Confusion Matrix into a 2 × 2 version enabling the use of the relevant performance metrics and methods like the Receiver Operating Characteristic and the Area Under the Curve for the assessment of different classification algorithms. The reduction method is based on class grouping and leads to a specific Confusion Matrix format. The developed method is then exploited for the assessment of several state-of-the-art machine learning algorithms applied on a customer experience metric called Net Promoter Score (NPS).

Multi-Class Confusion Matrix Reduction method and its application on Net Promoter Score classification problem

Man overboard incidents in a maritime vessel are serious accidents where, the efficient and rapid detection is crucial in the recovery of the victim. The severity of such accidents, urge the use of intelligent systems that are able to automatically detect a fall and provide relevant alerts. To this end the use of novel deep learning and computer vision algorithms have been tested and proved efficient in problems with similar structure. This paper presents the use of a deep learning framework for automatic detection of man overboard incidents. We investigate the use of simple RGB video streams for extracting specific properties of the scene, such as movement and saliency, and use convolutional spatiotemporal autoencoders to model the normal conditions and identify anomalies. Moreover, in this work we present a dataset that was created to train and test the efficacy of our approach.

Man Overboard: Fall detection using spatiotemporal convolutional autoencoders in maritime environments

Bioinformatics systems for monitoring and mitigating epidemics: The STAMINA Paradigm

Emotion recognition is a field of psychology that involves the process of identifying emotions and treating mental conditions like autism. The advancements in the field of machine learning and deep learning have paved the way for scientists to develop models for evaluating emotions by analyzing facial expressions, speech and text. However, the task of evaluating emotions could be best done by processing the bio-signals and neural imaging of the brain. In that sense, bio-signals such as Electroencephalogram (EEG) are less expensive to use and non-invasive, giving them an edge over traditional methods like Magnetic Resonant Imaging (MRI). However,
not many datasets are publicly available due to privacy issues and their availability is highly limited by the classification task. These constraints, along with the problem of data scarcity, motivates this work as an attempt to enhance the accuracy scores by generating synthetic features that are close to actual data distribution. In this research, we propose a Wasserstein Generative Adversarial Network with gradient penalty (WGAN-GP) based model that can help tackle this problem. The dataset that is investigated is DEAP, one of the benchmark datasets for evaluating emotion recognition algorithms. In the method proposed, nine descriptive features are extracted from the original data and baseline models are evaluated. 
Subsequently, a WGAN-GP is trained on these extracted features and is used to generate a new set of synthetic data features. The synthetic features are then analysed for quality and appended to the original data to expand this dataset. Experiments with different augmentation factors (x2, x3, x4) are investigated to evaluate the impact of the data augmentation procedure. The experimental results
demonstrate that the proposed method gives a considerable enhancement of the classification task’s performance.

Towards Real-time Generalized Ergonomic Risk Assessment for the Prevention of Musculoskeletal Disorders

Downloads

Next from PETRA 2021

Assistive Technologies for Supporting Wellbeing of Older Adults

Similar lecture

A Deep Learning Approach to Recognize Cognitive Load using PPG Signalshttps://dl.acm.org/doi/10.1145/3453892.3461625

Towards Real-time Generalized Ergonomic Risk Assessment for the Prevention of Musculoskeletal Disorders

.css-70qvj9{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;}Downloads

Next from PETRA 2021

Assistive Technologies for Supporting Wellbeing of Older Adults

Similar lecture

A Deep Learning Approach to Recognize Cognitive Load using PPG Signalshttps://dl.acm.org/doi/10.1145/3453892.3461625

Downloads