Mathematical Study of Nonlinear Mixed Convection Unsteady Flow in a Parallel Plate Inclined Channel in the Proximity of Time Periodic Boundary Conditions: Flow Reversal

This report is executed to examine the task of assimilating parameters on bipartite convection stream structure in a sloped pipeline while certain plate is disorderly warmed. The dictating motivation and energy identifications are ascertained and consequent expressions for thermal reading, liquid movement, fanning friction and stress flatten are acquired. The purpose of non-linear Boussinesq simulation is to escalate liquid movement, inverse stream generation at the channel plates, stress flatten, and fanning factor. In particular, the liquid motion escalates at the channel left portion and depletes at the channel right portion with the progress of time. A particular case of our development shows an excellent compromise with the previous consequences in the literature.

Contourlet-CNN for SAR Image Despeckling

A multiscale and multidirectional network named the Contourlet convolutional neural network (CCNN) is proposed for synthetic aperture radar (SAR) image despeckling. SAR image resolution is not higher than that of optical images. If the network depth is increased blindly, the SAR image detail information flow will become quite weak, resulting in severe vanishing/exploding gradients. In this paper, a multiscale and multidirectional convolutional neural network is constructed, in which a single-stream structure of convolutional layers is replaced with a multiple-stream structure to extract image features with multidirectional and multiscale properties, thus significantly improving the despeckling performance. With the help of the Contourlet, the CCNN is designed with multiple independent subnetworks to respectively capture abstract features of an image in a certain frequency and direction band. The CCNN can increase the number of convolutional layers by increasing the number of subnetworks, which makes the CCNN not only have enough convolutional layers to capture the SAR image features, but also overcome the problem of vanishing/exploding gradients caused by deepening the networks. Extensive quantitative and qualitative evaluations of synthetic and real SAR images show the superiority of our proposed method over the state-of-the-art speckle reduction method.

Unsupervised Deep Learning-Based RGB-D Visual Odometry

Recently, deep learning frameworks have been deployed in visual odometry systems and achieved comparable results to traditional feature matching based systems. However, most deep learning-based frameworks inevitably need labeled data as ground truth for training. On the other hand, monocular odometry systems are incapable of restoring absolute scale. External or prior information has to be introduced for scale recovery. To solve these problems, we present a novel deep learning-based RGB-D visual odometry system. Our two main contributions are: (i) during network training and pose estimation, the depth images are fed into the network to form a dual-stream structure with the RGB images, and a dual-stream deep neural network is proposed. (ii) the system adopts an unsupervised end-to-end training method, thus the labor-intensive data labeling task is not required. We have tested our system on the KITTI dataset, and results show that the proposed RGB-D Visual Odometry (VO) system has obvious advantages over other state-of-the-art systems in terms of both translation and rotation errors.

Enhanced Action Recognition Using Multiple Stream Deep Learning with Optical Flow and Weighted Sum

Various action recognition approaches have recently been proposed with the aid of three-dimensional (3D) convolution and a multiple stream structure. However, existing methods are sensitive to background and optical flow noise, which prevents from learning the main object in a video frame. Furthermore, they cannot reflect the accuracy of each stream in the process of combining multiple streams. In this paper, we present a novel action recognition method that improves the existing method using optical flow and a multi-stream structure. The proposed method consists of two parts: (i) optical flow enhancement process using image segmentation and (ii) score fusion process by applying weighted sum of the accuracy. The enhancement process can help the network to efficiently analyze the flow information of the main object in the optical flow frame, thereby improving accuracy. A different accuracy of each stream can be reflected to the fused score while using the proposed score fusion method. We achieved an accuracy of 98.2% on UCF-101 and 82.4% on HMDB-51. The proposed method outperformed many state-of-the-art methods without changing the network structure and it is expected to be easily applied to other networks.

Optimal In-Stream Structure Design through Considering Nitrogen Removal in Hyporheic Zone

The hyporheic zone (HZ), the region beneath or alongside a streambed, can play a vital role in a stream ecosystem. Previous studies have examined the impacts of in-stream structures on the HZ and river restoration; however, studies on optimizing the design of in-stream structures are still lacking. Therefore, this study aims to propose a method for optimizing the design of in-stream structures (e.g., weirs) through comprehensively considering both nitrogen removal amount (NRA) and nitrogen removal ratio (NRR) in the HZ based on numerical modelling. The Hydrologic Engineering Center’s River Analysis System (HEC-RAS) and COMSOL Multiphysics are employed for surface water and hyporheic flow simulations, respectively, and these two models are coupled by the hydraulic head along the surface of the streambed. The NRA and NRR are both closely related with residence time (RT), while the NRA is also influenced by hyporheic flux. Using the model outputs under different scenarios, regression equations for estimating the relevant variables (e.g., the maximum upstream distance in the subsurface flow influenced by the weir, the RT, and the hyporheic flux) are proposed. Then, the cumulative NRA (CNRA) and NRR can be calculated, and an objective function is formulated as the product of the normalized CNRA and NRR. The results show that the optimal height of the weir can be obtained based on the proposed method, and the validation shows the good general performance of this method. Sensitivity analysis indicates that the optimal height generally can be sensitive to the river discharge, i.e., the optimal height increases when the river discharge increases and vice versa. In addition, it is observed that, in the case of the optimal height, hyporheic flux increases when the slope increases while the influence of depth to bedrock on hyporheic flux is not significant. This study enhances our understanding of the optimal in-stream structure design, and potentially benefits river restoration in the face of continual degradation caused by human activities.

RDSNet: A New Deep Architecture forReciprocal Object Detection and Instance Segmentation

Object detection and instance segmentation are two fundamental computer vision tasks. They are closely correlated but their relations have not yet been fully explored in most previous work. This paper presents RDSNet, a novel deep architecture for reciprocal object detection and instance segmentation. To reciprocate these two tasks, we design a two-stream structure to learn features on both the object level (i.e., bounding boxes) and the pixel level (i.e., instance masks) jointly. Within this structure, information from the two streams is fused alternately, namely information on the object level introduces the awareness of instance and translation variance to the pixel level, and information on the pixel level refines the localization accuracy of objects on the object level in return. Specifically, a correlation module and a cropping module are proposed to yield instance masks, as well as a mask based boundary refinement module for more accurate bounding boxes. Extensive experimental analyses and comparisons on the COCO dataset demonstrate the effectiveness and efficiency of RDSNet. The source code is available at https://github.com/wangsr126/RDSNet.