RÉSUMÉ
Crowding occurs when the presence of nearby features causes highly visible objects to become unrecognizable. Although crowding has implications for many everyday tasks and the tremendous amounts of research reflect its importance, surprisingly little is known about how depth affects crowding. Most available studies show that stereoscopic disparity reduces crowding, indicating that crowding may be relatively unimportant in three-dimensional environments. However, most previous studies tested only small stereoscopic differences in depth in which disparity, defocus blur, and accommodation are inconsistent with the real world. Using a novel multi-depth plane display, this study investigated how large (0.54-2.25 diopters), real differences in target-flanker depth, representative of those experienced between many objects in the real world, affect crowding. Our findings show that large differences in target-flanker depth increased crowding in the majority of observers, contrary to previous work showing reduced crowding in the presence of small depth differences. Furthermore, when the target was at fixation depth, crowding was generally more pronounced when the flankers were behind the target as opposed to in front of it. However, when the flankers were at fixation depth, crowding was generally more pronounced when the target was behind the flankers. These findings suggest that crowding from clutter outside the limits of binocular fusion can still have a significant impact on object recognition and visual perception in the peripheral field.
While human eyesight is clearest at the point where the gaze is focused, peripheral vision makes objects to the side visible. This ability to detect movement and objects in a wider field of vision helps people to have a greater awareness of their surroundings. However, it is more difficult to identify an object using peripheral vision when it is surrounded by other items. This phenomenon is known as crowding and can affect many aspects of daily life, such as driving or spotting a friend in a crowd. In our three-dimensional world, peripheral objects are often at different distances. This variation in depth could influence the effect of crowding, yet little is known about its effect. While previous research has investigated the effect of small differences in depth on crowding, the studies did not replicate real-world conditions. To replicate depths that are likely to be encountered in the real world, Smithers et al. created a display using multiple screens positioned 0.4, 1.26 and 4 meters from the viewer. Images were displayed on the screens and researchers measured how well study participants could identify a target image when it was surrounded by similar, nearby images displayed closer or further away than the target. The experiments showed that most viewers are less able to recognize a target object when there are surrounding items and this effect is worsened when the items are separated from the object by large differences in depth. The findings show that instead of diminishing the effect of crowding as suggested by previous studies with small depth differences large depth differences that more closely recreate those encountered in the real world can amplify the effect of crowding. This greater understanding of how humans process objects in three-dimensional environments could help to better estimate the impact of crowding on people with eye and neurological disorders. In turn, the information could be used to design environments that are easier for such individuals to navigate.
Sujet(s)
Techniques histologiques , Perception visuelleRÉSUMÉ
The demand of microwave absorption materials (MAMs) with unique morphologies and electromagnetic (EM) balance has become necessary in recent years. Due to the ease of synthesis and tunable structure, metal-organic frameworks (MOFs) are widely used for this special MAMs. In this study, a new three-dimensional hybrid MOF is proposed that is co-doped with six equally branched star morphologies. The Co-C composite has the same six-branched morphology as that of the precursor. When the EM wave is incident, this special structure makes it easier for the EM wave to enter the material vertically due to the expansion of the incident surface, which is effective in adjusting the transmission path of the electron and the reflection and distribution of the EM wave. Because of the special morphology and magneto-dielectric synergy, the Co-C composite shows a minimum reflection loss (RLmin) of -48.5 dB at 11.0 GHz at an absorption thickness of 3.0 mm, with a microwave absorption bandwidth (EAB) of 6.1 GHz. This research provides a practical guidance for preparing the MAMs of special star structure.
RÉSUMÉ
Vehicle detection is an indispensable part of environmental perception technology for smart cars. Aiming at the issues that conventional vehicle detection can be easily restricted by environmental conditions and cannot have accuracy and real-time performance, this article proposes a front vehicle detection algorithm for smart car based on improved SSD model. Single shot multibox detector (SSD) is one of the current mainstream object detection frameworks based on deep learning. This work first briefly introduces the SSD network model and analyzes and summarizes its problems and shortcomings in vehicle detection. Then, targeted improvements are performed to the SSD network model, including major advancements to the basic structure of the SSD model, the use of weighted mask in network training, and enhancement to the loss function. Finally, vehicle detection experiments are carried out on the basis of the KITTI vision benchmark suite and self-made vehicle dataset to observe the algorithm performance in different complicated environments and weather conditions. The test results based on the KITTI dataset show that the mAP value reaches 92.18%, and the average processing time per frame is 15 ms. Compared with the existing deep learning-based detection methods, the proposed algorithm can obtain accuracy and real-time performance simultaneously. Meanwhile, the algorithm has excellent robustness and environmental adaptability for complicated traffic environments and anti-jamming capabilities for bad weather conditions. These factors are of great significance to ensure the accurate and efficient operation of smart cars in real traffic scenarios and are beneficial to vastly reduce the incidence of traffic accidents and fully protect people's lives and property.
RÉSUMÉ
Pedestrian detection is an important aspect of the development of intelligent vehicles. To address problems in which traditional pedestrian detection is susceptible to environmental factors and are unable to meet the requirements of accuracy in real time, this study proposes a pedestrian detection algorithm for intelligent vehicles in complex scenarios. YOLOv3 is one of the deep learning-based object detection algorithms with good performance at present. In this article, the basic principle of YOLOv3 is elaborated and analyzed firstly to determine its limitations in pedestrian detection. Then, on the basis of the original YOLOv3 network model, many improvements are made, including modifying grid cell size, adopting improved k-means clustering algorithm, improving multi-scale bounding box prediction based on receptive field, and using Soft-NMS algorithm. Finally, based on INRIA person and PASCAL VOC 2012 datasets, pedestrian detection experiments are conducted to test the performance of the algorithm in various complex scenarios. The experimental results show that the mean Average Precision (mAP) value reaches 90.42%, and the average processing time of each frame is 9.6 ms. Compared with other detection algorithms, the proposed algorithm exhibits accuracy and real-time performance together, good robustness and anti-interference ability in complex scenarios, strong generalization ability, high network stability, and detection accuracy and detection speed have been markedly improved. Such improvements are significant in protecting the road safety of pedestrians and reducing traffic accidents, and are conducive to ensuring the steady development of the technological level of intelligent vehicle driving assistance.