In this paper, we address the problem of people detection and tracking in crowded scenes using range cameras. We propose a new method for people detection and localisation based on the combination of background modelling and template matching. The method uses an adaptive background model in the range domain to characterise the scene without people. Then a 3D template is placed in possible people locations by projecting it in the background to reconstruct a range image that is most similar to the observed range image. We tested the method on a challenging outdoor dataset and compared it to two methods that each shares one characteristic with the proposed method: a similar template-based method that works in 2D and a well-known baseline method that works in the range domain. Our method performs significantly better, does not deteriorate in crowded environments and runs in real time.
People counting is a challenging task with many applications. We propose a method with a fixed stereo camera that is based on projecting a template onto the depth image. The method was tested on a challenging outdoor dataset with good results and runs in real time.
Large cities in the West respond to an ever-increasing shortage of affordable housing by accelerating the process of urban densification. Amsterdam, for instance, aims to increase its housing stock by 10 percent in the next 15 years as its population is expected to grow by 20 percent. As in other cities, it seems inevitable that high-rise buildings with higher skyscrapers than in the past will be built within the existing urban fabric. Such large-scale (re)development projects shape the conditions for inhabitants’ eye-level experiences, perception of place and overall well-being. The new hybrid field of neuroarchitecture offers promising eye-tracking technology and theories for measuring inhabitants’ visual experiences of the city and rethinking the effectiveness of applied design principles across the globe. In this paper, the ‘classic’ design solutions for creating streetscapes on a human scale in densified areas have been assessed by eye-tracking 31 participants in a laboratory setting, all of whom viewed photographs of 15 existing streetscapes in high-rise environments. The study drew on theories from the field of neuroarchitecture and used input from a panel of (landscape) architects and urban designers to design the research and analyze the eye-tracked patterns. The results indicate that the classic design principles (horizontal–vertical rhythms and variety; active ground floor; tactile materials) play a significant role in people’s appreciation of the streetscape and that their attention is unconsciously captured by the presence of these principles. The absence of the design principles seems to result in a scattered ‘searching’ eye movement pattern. This also suggests that a coherent design of streetscapes in high-rise environments may contribute to a human scale at eye-level.
MULTIFILE
Nederland kent ongeveer 220.000 bedrijfsongevallen per jaar (met 60 mensen die overlijden). Vandaar dat elke werkgever verplicht is om bedrijfshulpverlening (BHV) te organiseren, waaronder BHV-trainingen. Desondanks brengt slechts een-derde van alle bedrijven de arbeidsrisico’s in kaart via een Risico-Inventarisatie & Evaluatie (RI&E) en blijft het aandeel werknemers met een arbeidsongeval hoog. Daarom wordt er continu geïnnoveerd om BHV-trainingen te optimaliseren, o.a. door middel van Virtual Reality (VR). VR is niet nieuw, maar is wel doorontwikkeld en betaalbaarder geworden. VR biedt de mogelijkheid om veilige realistische BHV-noodsimulaties te ontwikkelen waarbij de cursist het gevoel heeft daar echt te zijn. Ondanks de toename in VR-BHV-trainingen, is er weinig onderzoek gedaan naar het effect van VR in BHV-trainingen en zijn resultaten tegenstrijdig. Daarnaast zijn er nieuwe technologische ontwikkelingen die het mogelijk maken om kijkgedrag te meten in VR m.b.v. Eye-Tracking. Tijdens een BHV-training kan met Eye-Tracking gemeten worden hoe een instructie wordt opgevolgd, of cursisten worden afgeleid en belangrijke elementen (gevaar en oplossingen) waarnemen tijdens de simulatie. Echter, een BHV-training met VR en Eye-Tracking (interacties) bestaat niet. In dit project wordt een prototype ontwikkeld waarin Eye-Tracking wordt verwerkt in een 2021 ontwikkelde VR-BHV-training, waarin noodsituaties zoals een kantoorbrand worden gesimuleerd (de BHVR-toepassing). Door middel van een experiment zal het prototype getest worden om zo voor een deel de vraag te beantwoorden in hoeverre en op welke manier Eye-Tracking in VR een meerwaarde biedt voor (RI&E) BHV-trainingen. Dit project sluit daarmee aan op het missie-gedreven innovatiebeleid ‘De Veiligheidsprofessional’ en helpt het MKB dat vaak middelen en kennis ontbreekt voor onderzoek naar effectiviteit rondom innovatieve-technologieën in educatie/training. Het project levert onder meer een prototype op, een productie-rapport en onderzoeks-artikel, en staat open voor nieuwe deelnemers bij het schrijven van een grotere aanvraag rondom de toepassing en effect van VR en Eye-Tracking in BHV-trainingen.
The Netherlands has approximately 220,000 industrial accidents per year (with 60 people who die). That is why every employer is obliged to organize company emergency response (BHV), including emergency response training. Despite this, only one-third of all companies map out their occupational risks via a Risk Inventory & Evaluation (RI&E) and the share of employees with an occupational accident remains high. That is why there is continuous innovation to optimize emergency response training, for example by means of Virtual Reality (VR). VR is not new, but it has evolved and become more affordable. VR offers the possibility to develop safe realistic emergency response simulations where the student has the feeling that they are really there. Despite the increase in VR-BHV training, little research has been done on the effect of VR in ER training and results are contradictory. In addition, there are new technological developments that make it possible to measure viewing behavior in VR using Eye-Tracking. During an emergency response training, Eye-Tracking can be used to measure how an instruction is followed, whether students are distracted and observe important elements (danger and solutions) during the simulation. However, emergency response training with VR and Eye-Tracking (interactions) does not exist. In this project, a prototype is being developed in which Eye-Tracking is incorporated into a VR-BHV training that was developed in 2021, in which emergency situations such as an office fire are simulated (the BHVR application). The prototype will be tested by means of an experiment in order to partly answer the question to what extent and in what way Eye-Tracking in VR offers added value for (RI&E) emergency response training. This project is therefore in line with the mission-driven innovation policy 'The Safety Professional' and helps SMEs that often lack resources and knowledge for research into the effectiveness of innovative technologies in education/training. The project will include a prototype, a production report and research article, and is open to new participants when writing a larger application about the application and effect of VR and Eye-Tracking in emergency response training.
A world where technology is ubiquitous and embedded in our daily lives is becoming increasingly likely. To prepare our students to live and work in such a future, we propose to turn Saxion’s Epy-Drost building into a living lab environment. This will entail setting up and drafting the proper infrastructure and agreements to collect people’s location and building data (e.g. temperature, humidity) in Epy-Drost, and making the data appropriately available to student and research projects within Saxion. With regards to this project’s effect on education, we envision the proposal of several derived student projects which will provide students the opportunity to work with huge amounts of data and state-of-the-art natural interaction interfaces. Through these projects, students will acquire skills and knowledge that are necessary in the current and future labor-market, as well as get experience in working with topics of great importance now and in the near future. This is not only aligned with the Creative Media and Game Technologies (CMGT) study program’s new vision and focus on interactive technology, but also with many other education programs within Saxion. In terms of research, the candidate Postdoc will study if and how the data, together with the building’s infrastructure, can be leveraged to promote healthy behavior through playful strategies. In other words, whether we can persuade people in the building to be more physically active and engage more in social interactions through data-based gamification and building actuation. This fits very well with the Ambient Intelligence (AmI) research group’s agenda in Augmented Interaction, and CMGT’s User Experience line. Overall, this project will help spark and solidify lasting collaboration links between AmI and CMGT, give body to AmI’s new Augmented Interaction line, and increase Saxion’s level of education through the dissemination of knowledge between researchers, teachers and students.
