Monocular Pedestrian Detection: Survey and Experiments

Pedestrian detection is a rapidly evolving area in computer vision with key applications in intelligent vehicles, surveillance, and advanced robotics. The objective of this paper is to provide an overview of the current state of the art from both methodological and experimental perspectives. The first part of the paper consists of a survey. We cover the main components of a pedestrian detection system and the underlying models. The second (and larger) part of the paper contains a corresponding experimental study. We consider a diverse set of state-of-the-art systems: wavelet-based AdaBoost cascade [74], HOG/linSVM [11], NN/LRF [75], and combined shape-texture detection [23]. Experiments are performed on an extensive data set captured onboard a vehicle driving through urban environment. The data set includes many thousands of training samples as well as a 27-minute test sequence involving more than 20,000 images with annotated pedestrian locations. We consider a generic evaluation setting and one specific to pedestrian detection onboard a vehicle. Results indicate a clear advantage of HOG/linSVM at higher image resolutions and lower processing speeds, and a superiority of the wavelet-based AdaBoost cascade approach at lower image resolutions and (near) real-time processing speeds. The data set (8.5 GB) is made public for benchmarking purposes.     

Similarity-based prioritization of test case automation

The importance of efficient software testing procedures is driven by an ever increasing system complexity as well as global competition. In the particular case of manual test cases at the system integration level, where thousands of test cases may be executed before release, time must be well spent in order to test the system as completely and as efficiently as possible. Automating a subset of the manual test cases, i.e, translating the manual instructions to automatically executable code, is one way of decreasing the test effort. It is further common that test cases exhibit similarities, which can be exploited through reuse when automating a test suite. In this paper, we investigate the potential for reducing test effort by ordering the test cases before such automation, given that we can reuse already automated parts of test cases. In our analysis, we investigate several approaches for prioritization in a case study at a large Swedish vehicular manufacturer. The study analyzes the effects with respect to test effort, on four projects with a total of 3919 integration test cases constituting 35,180 test steps, written in natural language. The results show that for the four projects considered, the difference in expected manual effort between the best and the worst order found is on average 12 percentage points. The results also show that our proposed prioritization method is nearly as good as more resource demanding meta-heuristic approaches at a fraction of the computational time. Based on our results, we conclude that the order of automation is important when the set of test cases contain similar steps (instructions) that cannot be removed, but are possible to reuse. More precisely, the order is important with respect to how quickly the manual test execution effort decreases for a set of test cases that are being automated.     

Measurement of porcine haptoglobin in meat juice using surface acoustic wave biosensor technology

Aim of the study was the application of biosensor technique to measure the concentration of an acute phase protein (APP) within complex matrices from animal origin. For the first time, acute phase protein haptoglobin (Hp) was detected from unpurified meat juice of slaughter pigs by a label-free biosensor-system, the SAW-based sam®5 system. The system uses a sensor chip with specific antibodies to catch Hp while the mass-related phase shift is measured. The concentration is calculated as a function of these measured phase shifts. The results correlate very well with reference measurement results obtained by enzyme-linked immunosorbent assay (ELISA), R = 0.98. The robust setup of the surface acoustic wave (SAW)-based system and its ability to measure within very short time periods qualifies it for large-scale analyses and is apt to identify rapidly pigs in the meat production process whose consumption would have an increased risk for consumers.     

Ad hoc self-healing of OFDMA networks using UAV-based relays

Due to the demand for constantly increasing data rates, especially within LTE (Long Term Evolution) and LTE-Advanced networks, today’s mobile radio systems are operating more and more at their capacity limits. Consequently, techniques to compensate cell outage and cell overload situations are needed. Existing approaches to address these problems are typically static solutions and are accompanied by Inter-Cell Interference (ICI), like the introduction of femtocells. In this paper we propose an UAV-based (Unmanned Aerial Vehicle) approach in order to provide ad hoc self-healing of cellular networks. Thereby, the UAVs are acting as mobile relays and offload traffic to surrounding cells with free resources. We introduce the genetic Interference-aware Positioning of Aerial Relays (IPAR) algorithm which is able to find suitable positions for the UAVs that maximize the downlink throughput of the cellular network. The algorithm is evaluated for hexagonal as well as irregular cellular network layouts and is based on an analytical model for deriving the key performance indicators throughput, system loss and free remaining resources in the cells for the downlink. In order to decrease the introduced interference between the UAVs and surrounding base stations, a random frequency allocation scheme is applied. Based on an exact analytical model for the SINR (Signal to Interference and Noise Ratio) of users served by UAVs, it is shown that by different hopping distributions a reduction of the SINR up to 11.6 dB can be achieved compared to the worst case performance of systems without resource planning.     

Self reported risk taking and risk compensation in skiers and snowboarders are associated with sensation seeking

In alpine skiing, a controversial discussion has been taking place regarding the potential influence of wearing a ski helmet on the individual level of risk taking behaviour. The aim of this study was to evaluate whether self reported risk taking behaviour and self reported risk compensation are associated with the personality trait sensation seeking (SS) in alpine skiing and snowboarding. In total, 683 persons (36% males and 64% females) completed an online-survey about attitudes and use of protective gear in winter sports including the German version of the sensation seeking scale form V. A logistic regression analysis including gender, age, nationality, preferred winter sport, self reported skiing ability, mean skiing time per season, use of ski helmets, and SS total score was used to estimated adjusted odds ratios (ORs) and their 95% confidence intervals (95 CI) for self reported risk taking behaviour. Regression analysis revealed that a more risky behaviour increased with male gender (OR: 2.7), with an age < 25 years (OR: 1.6), with skiing (OR: 1.3), higher skill level (OR: 5.7), and a mean skiing time > 28 days per season (OR: 2.2). In addition, SS total score was significantly higher in more risky compared to more cautious people (23.8 vs. 20.3, p < .001). Ski helmet use was not found to be predictive for a more risky behaviour (p > .05). Also, skiers and snowboarders with self reported risk compensation while wearing a ski helmet had higher SS total scores compared to those who did not report risk compensation (23.8 vs. 20.9, p = .001). In addition, self reported risk compensation in helmet wearers increased with an age < 25 years (OR: 2.2), a higher skill level (OR: 2.5) and a mean skiing time > 28 days per season (OR: 2.1). In conclusion, self reported risk taking and self reported risk compensation are associated with higher sensation seeking total scores. The personality trait sensation seeking, not wearing of a ski helmet, appears to be associated with riskier behaviour on the ski slopes.     

Substrate Templating upon Self‐Assembly of Hydrogen‐Bonded Molecular Networks on an Insulating Surface

Molecular self‐assembly on insulating surfaces, despite being highly relvant to many applications, generally suffers from the weak molecule–surface interactions present on dielectric surfaces, especially when benchmarked against metallic substrates. Therefore, to fully exploit the potential of molecular self‐assembly, increasing the influence of the substrate constitutes an essential prerequisite. Upon deposition of terephthalic acid and trimesic acid onto the natural cleavage plane of calcite, extended hydrogen‐bonded networks are formed, which wet the substrate. The observed structural complexity matches the variety realized on metal surfaces. A detailed analysis of the molecular structures observed on calcite reveals a significant influence of the underlying substrate, clearly indicating a substantial templating effect of the surface on the resulting molecular networks. This work demonstrates that choosing suitable molecule/substrate systems allows for tuning the balance between intermolecular and molecule–surface interactions even in the case of typically weakly interacting insulating surfaces. This study, thus, provides a strategy for deliberately exploiting substrate templating to increase the structural variety in molecular self‐assembly on a bulk insulator at room temperature.     

Dynamic saturation curve measurement for resin flow in glass fibre reinforcement

A methodology is presented to determine the saturation curve of a resin/glass fabric system, during infiltration in a transparent mould under constant flow rate. Video acquisitions are transformed by image analysis into saturation level versus position and time, and coupled to inlet pressure measurements. A numerical multiphase flow model is then used to simulate the infiltration for various combinations of drainage curve parameters. The numerical parameters to describe the saturation and relative permeability are determined by response surface optimization. The drainage curve and relative permeability equations determined at one time are shown to adequately describe the entire injection process, and to be flow-rate dependent.