Detecting vehicles in aerial imagery plays an important role in a wide range of applications. The current vehicle detection methods are mostly based on sliding-window search and handcrafted or shallow-learning-based features, having limited description capability and heavy computational costs. Recently, due to the powerful feature representations, region convolutional neural networks (CNN) based detection methods have achieved state-of-the-art performance in computer vision, especially Faster R-CNN. However, directly using it for vehicle detection in aerial images has many limitations: (1) region proposal network (RPN) in Faster R-CNN has poor performance for accurately locating small-sized vehicles, due to the relatively coarse feature maps; and (2) the classifier after RPN cannot distinguish vehicles and complex backgrounds well. In this study, an improved detection method based on Faster R-CNN is proposed in order to accomplish the two challenges mentioned above. Firstly, to improve the recall, we employ a hyper region proposal network (HRPN) to extract vehicle-like targets with a combination of hierarchical feature maps. Then, we replace the classifier after RPN by a cascade of boosted classifiers to verify the candidate regions, aiming at reducing false detection by negative example mining. We evaluate our method on the Munich vehicle dataset and the collected vehicle dataset, with improvements in accuracy and robustness compared to existing methods.

No abstract

No abstract

Earlier studies have suggested abnormal brain volumes in autism, but inconsistencies exist. Using voxel-based morphometry, we compared global and regional brain volumes in 17 high-functioning autistic children with 15 matched controls. We identified significant reduction in left white matter volume and white/gray matter ratio in autism. Regional brain volume reductions were detected for right anterior cingulate, left superior parietal lobule white matter volumes, and right parahippocampal gyrus gray matter volume, whereas enlargements in bilateral supramarginal gyrus, right postcentral gyrus, right medial frontal gyrus, and right posterior lobe of cerebellum gray matter in autism. Our findings showed global and regional brain volumes abnormality in high-functioning autism.

Vehicle detection with orientation estimation in aerial images has received widespread interest as it is important for intelligent traffic management. This is a challenging task, not only because of the complex background and relatively small size of the target, but also the various orientations of vehicles in aerial images captured from the top view. The existing methods for oriented vehicle detection need several post-processing steps to generate final detection results with orientation, which are not efficient enough. Moreover, they can only get discrete orientation information for each target. In this paper, we present an end-to-end single convolutional neural network to generate arbitrarily-oriented detection results directly. Our approach, named Oriented_SSD (Single Shot MultiBox Detector, SSD), uses a set of default boxes with various scales on each feature map location to produce detection bounding boxes. Meanwhile, offsets are predicted for each default box to better match the object shape, which contain the angle parameter for oriented bounding boxes' generation. Evaluation results on the public DLR Vehicle Aerial dataset and Vehicle Detection in Aerial Imagery (VEDAI) dataset demonstrate that our method can detect both the location and orientation of the vehicle with high accuracy and fast speed. For test images in the DLR Vehicle Aerial dataset with a size of 5616 × 3744, our method achieves 76.1% average precision (AP) and 78.7% correct direction classification at 5.17 s on an NVIDIA GTX-1060.

No abstract

The rules governing cerebellar output are not fully understood, but must involve Purkinje cell (PC) activity, as PCs are the major input to deep cerebellar nuclear (DCN) cells (which form the majority of cerebellar output). Here, the influence of PC complex spikes (CSs) was investigated by simultaneously recording DCN activity with CSs from PC arrays in anesthetized rats. Crosscorrelograms were used to identify PCs that were presynaptic to recorded DCN cells (presynaptic PCs). Such PCs were located within rostrocaudal cortical strips and displayed synchronous CS activity. CS-associated modulation of DCN activity included a short-latency post-CS inhibition and long-latency excitations before and after the CS. The amplitudes of the post-CS responses correlated with the level of synchronization among presynaptic PCs. A temporal precision of ≤10 ms was generally required for CSs to be maximally effective. The results suggest that CS synchrony is a key control parameter of cerebellar output.Editorial note: This article has been through an editorial process in which the authors decide how to respond to the issues raised during peer review. The Reviewing Editor's assessment is that all the issues have been addressed (see decision letter).

Key points Purkinje cells are the sole output of the cerebellar cortex and fire two distinct types of action potential: simple spikes and complex spikes.Previous studies have mainly considered complex spikes as unitary events, even though the waveform is composed of varying numbers of spikelets.The extent to which differences in spikelet number affect simple spike activity (and vice versa) remains unclear.We found that complex spikes with greater numbers of spikelets are preceded by higher simple spike firing rates but, following the complex spike, simple spikes are reduced in a manner that is graded with spikelet number.This dynamic interaction has important implications for cerebellar information processing, and suggests that complex spike spikelet number may maintain Purkinje cells within their operational range. AbstractPurkinje cells are central to cerebellar function because they form the sole output of the cerebellar cortex. They exhibit two distinct types of action potential: simple spikes and complex spikes. It is widely accepted that interaction between these two types of impulse is central to cerebellar cortical information processing. Previous investigations of the interactions between simple spikes and complex spikes have mainly considered complex spikes as unitary events. However, complex spikes are composed of an initial large spike followed by a number of secondary components, termed spikelets. The number of spikelets within individual complex spikes is highly variable and the extent to which differences in complex spike spikelet number affects simple spike activity (and vice versa) remains poorly understood. In anaesthetized adult rats, we have found that Purkinje cells recorded from the posterior lobe vermis and hemisphere have high simple spike firing frequencies that precede complex spikes with greater numbers of spikelets. This finding was also evident in a small sample of Purkinje cells recorded from the posterior lobe hemisphere in awake cats. In addition, complex spikes with a greater number of spikelets were associated with a subsequent reduction in simple spike firing rate. We therefore suggest that one important function of spikelets is the modulation of Purkinje cell simple spike firing frequency, which has implications for controlling cerebellar cortical output and motor learning.