In this paper, we present a comprehensive review of the imbalance problems in object detection. To analyze the problems in a systematic manner, we introduce a problem-based taxonomy. Following this taxonomy, we discuss each problem in depth and present a unifying yet critical perspective on the solutions in the literature. In addition, we identify major open issues regarding the existing imbalance problems as well as imbalance problems that have not been discussed before. Moreover, in order to keep our review up to date, we provide an accompanying webpage which catalogs papers addressing imbalance problems, according to our problem-based taxonomy. Researchers can track newer studies on this webpage available at: https://github.com/kemaloksuz/ObjectDetectionImbalance.

Computational modeling of the primate visual system yields insights of potential relevance to some of the challenges that computer vision is facing, such as object recognition and categorization, motion detection and activity recognition, or vision-based navigation and manipulation. This paper reviews some functional principles and structures that are generally thought to underlie the primate visual cortex, and attempts to extract biological principles that could further advance computer vision research. Organized for a computer vision audience, we present functional principles of the processing hierarchies present in the primate visual system considering recent discoveries in neurophysiology. The hierarchical processing in the primate visual system is characterized by a sequence of different levels of processing (on the order of 10) that constitute a deep hierarchy in contrast to the flat vision architectures predominantly used in today's mainstream computer vision. We hope that the functional description of the deep hierarchies realized in the primate visual system provides valuable insights for the design of computer vision algorithms, fostering increasingly productive interaction between biological and computer vision research.

Drone detection is the problem of finding the smallest rectangle that encloses the drone(s) in a video sequence. In this study, we propose a solution using an end-to-end object detection model based on convolutional neural networks. To solve the scarce data problem for training the network, we propose an algorithm for creating an extensive artificial dataset by combining background-subtracted real images. With this approach, we can achieve precision and recall values both of which are high at the same time.

Detection and distance estimation of micro unmanned aerial vehicles (mUAVs) is crucial for (i) the detection of intruder mUAVs in protected environments; (ii) sense and avoid purposes on mUAVs or on other aerial vehicles and (iii) multi-mUAV control scenarios, such as environmental monitoring, surveillance and exploration. In this article, we evaluate vision algorithms as alternatives for detection and distance estimation of mUAVs, since other sensing modalities entail certain limitations on the environment or on the distance. For this purpose, we test Haar-like features, histogram of gradients (HOG) and local binary patterns (LBP) using cascades of boosted classifiers. Cascaded boosted classifiers allow fast processing by performing detection tests at multiple stages, where only candidates passing earlier simple stages are processed at the preceding more complex stages. We also integrate a distance estimation method with our system utilizing geometric cues with support vector regressors. We evaluated each method on indoor and outdoor videos that are collected in a systematic way and also on videos having motion blur. Our experiments show that, using boosted cascaded classifiers with LBP, near real-time detection and distance estimation of mUAVs are possible in about 60 ms indoors (1032 × 778 resolution) and 150 ms outdoors (1280 × 720 resolution) per frame, with a detection rate of 0.96 F-score. However, the cascaded classifiers using Haar-like features lead to better distance estimation since they can position the bounding boxes on mUAVs more accurately. On the other hand, our time analysis yields that the cascaded classifiers using HOG train and run faster than the other algorithms.