United States

We develop deep learning-based classifiers for Audio Event Detec-tion (AED), attacking them next with some white noise disturbances. We show that an attacker can use such simple disturbances to poten-tially fully avoid detection by AED systems. Prior work has shown that attackers can mislead image classification tasks, however this work focuses on attacks against AED systems, by tampering the audio and not image. This work brings awareness to the designers and manufacturers of AED systems and devices, as these solutions are becoming more ubiquitous by the day.

PETRA 2021

Attacking Audio Event Detection Deep Learning Classifiers with White Noise

Naval Information Warfare Center Pacific is partnering with AFCEA San Diego chapter to host the workshop.

The NAML workshop is an opportunity for researchers from Navy research and development centers and our partner organizations to exchange technical ideas and to come together as a team.

The goals of NAML are:

* To spread awareness of current machine learning research relevant to Naval applications

* To connect machine learning researchers with experts in military requirements

* To build and strengthen collaborations within the defense research community

**Watch this video to hear about the AFCEA purpose.**

<iframe width=1280 height=706 frameborder="0" scrolling="no" src="https://screencast-o-matic.com/player/creleVVcvRh?v=6&ff=1&title=0&controls=1" allowfullscreen="true"></iframe>

NAML 2021

demo

**About**
<br>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.

<br>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.

<br>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
<br>
<br>

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.

Hand pose estimation is getting a lot of attention in many areas such
as Human-Computer Interaction and Sign Language Recognition.
A fundamental step to accurately estimate the hand pose involves
detecting and localizing fingertips in an image. Despite the progress
of 2-D hand pose estimation in recent studies, accurate and robust
detection and localization of fingertips still remains a challenging
task due to low resolution of a fingertip in images and varying
lightning condition.
Inspired by the progress of the Generative Adversarial Network
(GAN) and image-style transfer, we propose a two-stage pipeline to
accurately localize the fingertip position even in varying lighting
and severe self occlusion on depth images. The idea is to use a
Cycle-consistent Generative Adversarial Network (Cycle-GAN) to
apply unpaired image-to-image translation and generate a depth
image with colored predictions on the fingertips, wrist, and palm
given a real depth image. The model is trained in a semi-supervised
manner using a collection of images from source and target domains
that do not need to be related in anyway. Then, by applying color
segmentation techniques, we localize the center of each colored area
which results in finding the location of each fingertip along with
center of the wrist and the palm. The proposed method achieves
visually promising results on noisy depth images captured using the
Microsoft Kinect. Experiments on the challenging NYU hand dataset
have demonstrated that our approach not only generates plausible
samples, but also outperforms state-of-the-art approaches on 2-D
fingertip estimation by a significant margin even in the presence
of severe self-occlusion and varying lighting conditions. Moreover,
fingertips would be detected irrespective of user orientation using
this method.

A pipeline for Hand 2-D Keypoint Localization using Unpaired Image to Image Translation

The ageing population leads to an increase of people requiring long-term care. Assisting people when getting out of bed and fast reactions to falls can help to reduce costs and the risk of injury. We describe the possibility of detecting getting up behavior from a bed using different deep learning models and depth data as a proof of concept. The hereby used computer vision approach uses unobtrusive depth data in order to protect people’s privacy. We gather data from different subjects, postures, views and rooms, with which we then train the network. Both classification and object detection methods are able to reliably detect getting up behavior. Situations that could not be correctly classified were when a person was changing from lying to sitting or when the legs of the person were covered with a blanket. Our results show that using pretrained networks is the key contributor in training. We also demonstrate that convolutional neural networks are capable of extracting highlevel task-dependent features from depth data which can be utilized
in developing ambient intelligent systems. The practicality of the network can be adapted from getting up from a bed to sitting and walking inside the room, based on the purpose of the real-life application.

Privacy preserving getup detection

Teletherapeutic devices enable new possibilities to track the therapy process by gathering and visualizing training statistics. For the therapist this can mean a quicker but also a deeper look at the therapy results. For the patient, the right presentation of data can foster the motivation and adherence to the therapy. In an iterative user centered design process, requirements of data visualization were gathered and the developed solution was validated with health professionals. For the therapists, the most important part is to have a comprehensive overview of conducted exercises and training duration. For the patient, the visualization should be simple and motivative. Further validation, especially with patients is needed.

Telerehabilitation Platform REHA2030: Visualization of training statistics for patient and therapist

In this paper we present early results of a novel approach for human activity recognition, focusing on activities of daily living. Our approach is multimodal, i.e., it combines RGB, depth and skeletal data measurements. Moreover, it is based on a novel hybrid deep neural network architecture that combines a convolutional neural network, a long short term memory network and it is evaluated using a publicly available dataset.

A Deep Hybrid Architecture for Human Activity Recognitionhttps://dl.acm.org/doi/10.1145/3453892.3461006

The simulated environment presented aims to provide a method for preparing a training set for most of this deep network adaptation. The goal is to use a simulation-generated training set for most or all of the data needed to achieve the transfer of knowledge from the original network to the platform-specific

A Simulated Environment For Traversability Estimation Experiments In Field Robotic Applications

Virtual Reality (VR) is attracting more and more attention from academic research and practical application with the current availability of low-cost and end-user friendly devices. In terms of haptic (rather than visual) interaction, however, this technology is still in its infancy and there are few devices that are inexpensive and technologically simple to operate, respectively to procure. In this context, we present a concept of how haptic interaction with VR data gloves can also succeed by means of a commercially available web cam, sophisticated tracking software, and homemade low-cost hardware. All hardware and software components are to be obtained inexpensively or are open-source in order to achieve the greatest possible dissemination potential. With this work, we intend to provide an important trigger for future improvements and dissemination in terms of both technology and areas of application.

Build-and-Touch: A Low-Cost, DIY, Open-Source Approach Towards Touchable Virtual Reality

Sign language (SL) technologies including various deaf accessibility tools, dynamic sign language synthesis systems, machine learning based signing video recognition algorithms and sign language machine translation, are only some of the technologies requiring the existence of sign language annotated resources. SL resources are equally critical for the development of various end-user tools and services, which support everyday communication needs via SL.
The poster demonstrates how annotated SL lexical and corpus resources, created and maintained at ILSP/ATHENA RC, serve the needs of both Sign Language Technologies and Greek Sign Language (GSL) signers’ accessibility.


The Landscape of Accessibility Tools Requiring SL Resources

FutureMe: Negotiating Learning Goals with your Future Learning-Self Avatar

Nurses have to carry out a myriad of tasks, which is one of the main reasons for burnout. In this paper, a robotic Multitasking Intelligent Nurse Aid (MINA) is proposed to assist nurses with everyday tasks and tackle nurse burnout. MINA uses Simultaneous Localization And Mapping (SLAM) to map shelves in a storage room and pointcloud using an RGB-Depth camera. The barcode of the item is first detected using zbar library, and then the pointcloud is generated in ROS and saved accordingly. Preliminary experiments with respect to the creation of a pointcloud map with barcode detection are presented. The preliminary results reveal that the system shows a 100% barcode detection rate, and with each barcode detection a corresponding pointcloud map is generated and saved automatically without any intervention.

MINA: A Multitasking Intelligent Nurse Aid Robot

Assessment and prevention of work-related musculo-skeletal disorders (WMSD) and fatigue are considered common and critical tasks towards occupational safety and well-being in working environments. Indicators/risk factors for WMSD and physical/mental fatigue in the manufacturing industry include labor-intensive assembly works, repetitive tasks, prolonged, sub-optimal working postures and more. We briefly present our work on automatic, non-obtrusive detection of worker physical strain using visual data acquired by cameras and on automatic detection of worker fatigue based on heart rate (HR) data acquired by smartwatches.


Attacking Audio Event Detection Deep Learning Classifiers with White Noise

Downloads

Next from PETRA 2021

A pipeline for Hand 2-D Keypoint Localization using Unpaired Image to Image Translation

Similar lecture

Artificial Intelligence Applications to Cybersecurity (AI ATAC) Prize Challenges I&II

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES