The joint detection and embedding (JDE) paradigm has attracted considerable attention in multiobject tracking due to its effectiveness and efficiency. However, there are some issues that hinder further improvement in the performance of JDE-based methods. These issues include feature competition between detection and reidentification tasks, as well as the underutilization of detections in data association. Furthermore, the applicability of the JDE-based methods in multiscenario is limited because they focus on optimization in a single scenario. To tackle these issues, we rethink the JDE paradigm from two critical aspects, i.e., feature representation and data association. Based on this, a powerful JDE-based tracking method is proposed, named MMTrack. Specifically, a feature decoupling network is proposed to mitigate competition between detection and reidentification tasks. In addition, a hierarchical association strategy is designed to maximize object matching by considering nearly all detections. Finally, comprehensive experiments are conducted on public datasets covering multiscenario, which demonstrate the superiority of MMTrack.

Addressing on monocular visual odometry problem, this paper presents a novel end-to-end network for estimation of camera ego-motion. The network learns the latent space of optical flow (OF) and models sequential dynamics so that the motion estimation is constrained by the relations between sequential images. We compute the OF field of consecutive images and extract the latent OF representation in a self-encoding manner. A Recurrent Neural Network is then followed to examine the OF changes, i.e., to conduct sequential learning. The extracted sequential OF latent space is used to compute the regression of the 6-dimensional pose vector. Particularly, we separately train the encoder in an unsupervised manner. By this means, we avoid non-convergence during the training of the whole network and allow more generalized and effective feature representation. Substantial experiments have been conducted on KITTI and Malaga datasets, and the results demonstrate that our model outperforms most learning-based VO approaches.

Local binary pattern (LBP) is one of the most successful feature descriptors. However, LBP and its variants have not been as successful as other feature descriptors in video anomaly detection (VAD). This is because LBP and its variants are mainly designed for spatial texture analysis. Although the volume LBP (VLBP) and the LBP-three orthogonal planes (LPB-TOP) have the capability of describing dynamic texture, they are seldom used as descriptors for VAD because 1) both VLBP and LBP-TOP are more suitable for natural scenes with rich dynamic textures, but sensitive to noise in the scenes with less dynamic textures, 2) the combination of motion and appearance not only limits their capability of motion characterizing but also brings the irrelevant appearance information such as background, and 3) high dimensionality is another drawback. In this paper, a new variant of the LBP called the squirrel-cage LBP (SCLBP) is proposed for VAD. By imitating the structure of squirrel cage rotor, the proposed SCLBP can be regarded as a stretched LBP in temporal direction and has two distinct features: 1) it is computed at vector-wise, rather than at pixel-wise (i.e., for the central vector with respect to its surrounding parallel vectors), and 2) the sign between two vectors is determined by the angle-based thresholding scheme. The SCLBP can effectively encode the motion information and is insensitive to noise and irrelevant disturbances caused by dynamic background and illumination change. To the best of our knowledge, The SCLBP is the first variant of the LBP specially designed for motion characterizing. The SCLBP has great flexibility, extendibility, and low dimensionality (only one-third of the LBP-TOP descriptor). The effectiveness of the proposed SCLBP descriptor is demonstrated on different public data sets and compared with the other dominant descriptors and state-of-the-art approaches in VAD.

Existing anomaly detection (AD) approaches rely on various hand‐crafted representations to represent video data and can be costly. The choice or designing of hand‐crafted representation can be difficult when faced with a new dataset without prior knowledge. Motivated by feature learning, e.g. deep leaning and the ability to directly learn useful representations and model high‐level abstraction from raw data, the authors investigate the possibility of using a universal approach. The objective is learning data‐driven high‐level representation for the task of video AD without relying on hand‐crafted representation. A deep incremental slow feature analysis (D‐IncSFA) network is constructed and applied to directly learning progressively abstract and global high‐level representations from raw data sequence. The D‐IncSFA network has the functionalities of both feature extractor and anomaly detector that make AD completion in one step. The proposed approach can precisely detect global anomaly such as crowd panic. To detect local anomaly, a set of anomaly maps, produced from the network at different scales, is used. The proposed approach is universal and convenient, working well in different types of scenarios with little human intervention and low memory and computational requirements. The advantages are validated by conducting extensive experiments on different challenge datasets.

It is extremely time-consuming to restart a long-running simulation from the beginning when a failure occurs. Checkpointing is a viable solution that enables simulations to be resumed from the point of failure. We study three models to determine the optimal checkpoint interval between contiguous checkpoints so that the total execution time is minimized and we demonstrate that optimal checkpointing can facilitate self-optimizing. This study greatly advances our knowledge of and practice in optimizing long-running scientific simulations.

Presented in this paper is a novel method to calibrate the co-ordinate systems used by two separate sensor devices for the purposes of sensor fusion. In this example the sensors are a camera and a LIDAR device which are observing the same scene from different viewpoints. Using a synthetic set of corresponding 2D image co-ordinates and 3D LIDAR measurements as reference data the task of aligning re-projected measurements with reference measurements was posed as an optimisation problem. The objective of the optimisation is to find a set of calibration parameters (external offsets and internal camera parameters) which minimise the sum of squared errors between the reference image co-ordinates and the re-projected data. The re-projected data is obtained by transforming the reference LIDAR measurements using the calibration parameters and the errors are defined as the straight-line distance between each reference and re-projected pixel pair. Using the Nelder-Mead simplex search method calibration parameters were found in under a second such that the sum of squared errors across a data set of 200 points was less than 0.19 i.e. average error per pixel of 0.031px. The method finds both internal and external calibration factors and makes no assumptions about the model. Furthermore if a second optimisation pass is made the error can be reduced to almost zero using only 4 reference pairs assuming these points are selected correctly.

The problem of packing congruent spheres (i.e., copies of the same sph ere) in a bounded domain arises in many applications. In this paper, we present a new pack-and-shake scheme for packing congruent spheres in various bounded 2-D domains. Our packing scheme is based on a number of interesting ideas, such as a trimming and packing approach, optimal lattice packing under translation and/or rotation, shaking procedures, etc. Our packing algorithms have fairly low time complexities. In certain cases, they even run in nearly linear time. Our techniques can be easily generalized to congruent packing of other shapes of objects, and are readily extended to higher dimensional spaces. Applications of our packing algorithms to treatment planning of radiosurgery are discussed. Experimental results suggest that our algorithms produce reasonably dense packings.