Joint vehicle localization and categorization in high resolution aerial images can provide useful information for applications such as traffic flow structure analysis. To maintain sufficient features to recognize small-scaled vehicles, a regions with convolutional neural network features (R-CNN) -like detection structure is employed. In this setting, cascaded localization error can be averted by equally treating the negatives and differently typed positives as a multi-class classification task, but the problem of class-imbalance remains. To address this issue, a cost-effective network extension scheme is proposed. In it, the correlated convolution and connection costs during extension are reduced by feature map selection and bi-partite main-side network construction, which are realized with the assistance of a novel feature map class-importance measurement and a new class-imbalance sensitive main-side loss function. By using an image classification dataset established from a set of traditional real-colored aerial images with 0.13 m ground sampling distance which are taken from the height of 1000 m by an imaging system composed of non-metric cameras, the effectiveness of the proposed network extension is verified by comparing with its similarly shaped strong counter-parts. Experiments show an equivalent or better performance, while requiring the least parameter and memory overheads are required.

Visual tracking is a basic task in many applications. However, the heavy computation and low speed of many recent trackers limit their applications in some computing power restricted scenarios. On the other hand, the simple update scheme of most correlation filter-based trackers restricts their robustness during target deformation and occlusion. In this paper, we explore the update scheme of correlation filter-based trackers and propose an efficient and adaptive training sample update scheme. The training sample extracted in each frame is updated to the training set according to its distance between existing samples measured with a difference hashing algorithm or discarded according to tracking result reliability. In addition, we expand our new tracker to long-term tracking. On the basis of the proposed model updating mechanism, we propose a new tracking state discrimination mechanism to accurately judge tracking failure, and resume tracking after the target is recovered. Experiments on OTB-2015, Temple Color 128 and UAV123 (including UAV20L) demonstrate that our tracker performs favorably against state-of-the-art trackers with light computation and runs over 100 fps on desktop computer with Intel i7-8700 CPU(3.2 GHz).

The mean shift algorithm has been introduced successfully into the field of computer vision to be an efficient approach for visual tracking but the tracker has been awkward in handling the scale change of the object. This study addresses the scale estimation problem of the mean shift tracker, and proposes a novel method which is based on invariant foreground occupation ratio to solve this problem. The foreground occupation ratio is defined as the proportion of the foreground pixels in an image region. By taking an analysis of the foreground occupation ratio, the authors obtain its three simple properties. With its property of scale invariance, an iterative approximation approach is employed to estimate the scale of the foreground in the current image. The scale value is modified by a weighting function, and it is adjusted along the two axes with respect to the width and the height of the target. The scale estimation algorithm is then employed in the mean shift tracker to obtain the ability of scale adaptation for tracking. Experimental results show that, using the authors method for object scale estimation, the mean shift tracker performs well in tracking the target efficiently when its scale continuously changes.

Existing image classification methods based on convolutional neural networks usually use a large number of samples to learn classification features hierarchically, causing the problems of over-fitting and error propagation layer by layer. Thus, they are vulnerable to adversarial samples generated by adding imperceptible disturbances to input samples. To address the above issue, we propose a cognitive-driven color prior model to memorize the color attributes of target samples inspired by the characteristics of human memory. At inference stage, color priors are indexed from the memory and fused with features of convolutional neural networks to achieve robust image classification. The proposed color prior model is cognitive-driven and has no training parameters, thus it has strong generalization and can effectively defend against adversarial samples. In addition, our method directly combines the features of the prior model with the classification probability of the convolutional neural network, without changing the network structure and its parameters of the existing algorithm. It can be combined with other adversarial attack defense methods, such as various preprocessing modules such as PixelDefense or adversarial training methods, to improve the robustness of image classification. Experiments on several benchmark datasets show that the proposed method improves the anti-interference ability of image classification algorithms.