Eye-based keystroke prediction for natural texts - a feasibility analysis

The use of videoconferencing is on the rise after COVID-19, being common to look at the screen and see someone typing. A side-channel attack may be launched to infer the text written from the face image. In this paper, we analyse the feasibility of such an attack, being the first proposal which work with a complete keyset (50 keys) and natural texts. We use different scenarios, lighting conditions and natural texts to increase realism. Our study involves 30 participants, who typed 49,365 keystrokes. We characterize the effect of lighting, gender, age and use of glasses. Our results show that on average 13.71% of keystrokes are revealed without error, and up to 31.8%, 52.5% and 61.2% are guessed with a maximum error of 1, 2 and 3 keys, respectively. © 2022 IEEE.

Light exposure in home-based work: Can a simple lighting system increase alertness?

Since the proportion of digital and more flexible work in the western labour market increases, more and more employees are working at least partly from home. This development was even enhanced by the COVID-19-pandemic. In contrast to office workplaces, lighting at home-based workplaces is less studied and regulated. Lighting has been shown to not only ensure vision but also evoke non-image forming effects such as changes in alertness. In this study, light exposure of nine office employees at their home-based workplaces was investigated. Illuminance at home-based workplaces was found to be low, compared to office standards. In addition, melanopic equivalent daylight illuminance (MEDI) did not reach recommendations for healthy daytime light exposure. Furthermore, an additional lighting was installed at participants' desks in order to examine possible effects on alertness. Mean illuminance and MEDI during work were increased by the additional lighting. A decrease in subjective sleepiness could be shown after 6 hours, although differences were not significant. Improvements of response time in a psychomotor vigilance task were already achieved at the beginning of work and after 3 hours. This study shows that lighting conditions at home-based workplaces often do not meet the criteria for health-promoting lighting in terms of non-image forming effects. © The Chartered Institution of Building Services Engineers 2023.

Deep Learning for Crowd Image Classification for Images Captured Under Varying Climatic and Lighting Condition

Most of recent events have attracted a lot of attention towards importance of automatic crowd classification and management. COVID-19 is the most setback for the entire world. During these events proper breakout and public crowd management leads to the requirement of managing, counting, securing as well as tracking the crowd. But automatic analysis of the crowd is very challenging task because of varying climatic and lighting conditions, varying postures etc. During this paper we have developed PYTHON based system for automatic crowd images classification using Deep learning. This paper is the first attempt for automatic classification of crowd images. We have prepared the dataset of crowd classification consisting of three categories. The proposed methodology of crowd classification starts with preprocessing during which we have used median filtering for noise removal. Deep learning models are developed using 70% training images. The performance of the system is evaluated for various deep learning algorithms including one block VGG, two block VGG and three block VGG. We have also evaluated the performance of three block VGG using dropout. VGG16 transfer learning based crowd classification is developed using PYTHON. Using VGG16 transfer learning we achieved the accuracy of 69.44.% which is highest among all deep learning classification models during this study © 2022 IEEE.

‘Personas for lighting'. Three methods to develop personas for the indoor lighting environment

The objective of this research is to describe and compare three different methods of generating ‘persona for lighting' to envision users' behaviour within the lighting environment. ‘Personas' are used to represent typical users, highlighting their needs, perspectives, and expectations to aid user-centric design approaches. The researchers looked for the most useful method of shaping ‘personas for lighting' to learn about users' satisfaction with the various lighting conditions to identify their needs. Method one of lighting persona development, was based on interviews with 87 users of five buildings of four different types: an office, a primary school, two university buildings, and a factory. The lighting conditions were observed and measured in all the buildings. As a result, 22 personas for lighting were created. In method two personas were generated based on pre-interviews, workshops on lighting and post-interviews with ten users along with the onsite lighting measurements. Later, due to the Covid-19 pandemic's lockdowns, an online survey on the visual lighting environment in home offices was carried out among 694 students and professionals from seven countries to create two more personas for lighting (method three). All 26 ‘personas for lighting' were generated in relation to observed lighting conditions, based on the satisfaction, preferences and needs of the users working within variously lit indoor environments. All the tested methods can be used for nearly any type of building and room, but the resulting personas are different due to the specific limitations of the methods. The created personas may help to identify future users' lighting preferences, needs and requirements and assist designers. However, to fully understand their impact on the lighting research practice they should be tested in real projects. © 2022 The Author(s)

Automated Thermal Screening for COVID-19 Using Machine Learning

In the last two years, millions of lives have been lost due to COVID-19. Despite the vaccination programmes for a year, hospitalization rates and deaths are still high due to the new variants of COVID-19. Stringent guidelines and COVID-19 screening measures such as temperature check and mask check at all public places are helping reduce the spread of COVID-19. Visual inspections to ensure these screening measures can be taxing and erroneous. Automated inspection ensures an effective and accurate screening. Traditional approaches involve identification of faces and masks from visual camera images followed by extraction of temperature values from thermal imaging cameras. Use of visual imaging as a primary modality limits these applications only for good-lighting conditions. The use of thermal imaging alone for these screening measures makes the system invariant to illumination. However, lack of open source datasets is an issue to develop such systems. In this paper, we discuss our work on using machine learning over thermal video streams for face and mask detection and subsequent temperature screening in a passive non-invasive way that enables an effective automated COVID-19 screening method in public places. We open source our NTIC dataset that was used for training our models and was collected at 8 different locations. Our results show that the use of thermal imaging is as effective as visual imaging in the presence of high illumination. This performance stays the same for thermal images even under low-lighting conditions, whereas the performance with visual trained classifiers show more than 50% degradation. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Benzene and NOx photocatalytic-assisted removal using indoor lighting conditions

Modern life-style is creating an indoor generation: human beings spend approximately 90% of their time indoors, almost 70% of which is at home – this trend is now exacerbated by the lockdowns/restrictions imposed due to the COVID-19 pandemic. That large amount of time spent indoors may have negative consequences on health and well-being. Indeed, poor indoor air quality is linked to a condition known as sick building syndrome. Therefore, breathing the freshest air possible is of outmost importance. Still, due to reduced ventilation rates, indoor air quality can be considerably worse than outdoor. Heating, ventilation, and air conditioning (HVAC), air filtration systems and a well-ventilated space are a partial answer. However, these approaches involve only a physical removal. The photocatalytic mineralization of pollutants into non-hazardous, or at least less dangerous compounds, is a more viable solution for their removal. Titanium dioxide, the archetype photocatalytic material, needs UVA light to be ‘activated’. However, modern household light emitting diode lamps irradiate only in the visible region of the solar spectrum. We show that the surface of titanium dioxide nanoparticles modified with copper oxide(s) and graphene has promise as a viable way to remove gaseous pollutants (benzene and nitrogen oxides) using a common light emitting diode bulb, mimicking real indoor lighting conditions. Titanium dioxide, modified with 1 mol% CuxO and 1 wt% graphene, proved to have a stable photocatalytic degradation rate, three times higher than that of unmodified titania. Materials produced in this research work are thus strong candidates for offering a safer indoor environment. © 2022 Elsevier Ltd

Detection of Mask Wearing in Dim Light Based on Attention Mechanism

Since the outbreak of COVID-19, the detection of wearing masks has become a necessary measure for epidemic prevention and control. To solve the problem about low accuracy of mask wearing detection under dim lighting conditions, a method of mask wearing detection combining attention mechanism with YOLOv5 network model is proposed, which uses image enhancement algorithm to pre-process the training set pictures, and then put these pictures to YOLOv5 network with attention mechanism for iterative training. After training, the optimal weight is saved and the best model is used to test the accuracy on the test set. The experimental results show that the YOLOv5 network model with attention mechanism can effectively enhance the extraction of key points such as face and mask and improve the robustness of the model. The accuracy of mask wearing can reach 92% under dim lighting conditions, which can effectively meet the actual needs. Copyright ©2022 Journal of University of Electronic Science and Technology of China. All rights reserved.