Searcharxiv⌕ Search

arXiv subjects

Bin Zhao

Publications and source records attributed to Bin Zhao.

142 records · Page 8Linked to original sources
arXiv2020

Modeling and optimization of radiative cooling based thermoelectric generators

The possibility of night-time power generation has recently stimulated interest in using the radiative sky cooling mechanism with thermoelectric generators (TEG). These passive, low-temperature difference devices have been shown to generate electricity at night with no active input of heat needed, instead using the ambient air itself as the heat source. Here, we optimize both the geometry and operating conditions of radiative cooling driven thermoelectric (RC-TE) generators. We determine the optimal operating conditions, including maximum power point and maximum efficiency point, by developing a combined thermal and electrical model. Our results show that the optimal operating condition results in larger power output than was previously expected. Moreover, we show that maximum power density occurs when the area ratio between cooler and P or N element reaches an optimal value. Finally, we perform a parametric study that takes account of environmental and structural parameters to improve the performance of the RC-TE device, including enhancing heat transfer between the hot surface and ambient air, suppressing the cooling loss of the radiative cooler, and optimizing the geometry of individual thermocouples. Our work identifies how to maximize the output of RC-TE devices and provides comprehensive guidance on making use of this new passive power generation method.

physics.app-ph↗
arXiv2020

Spectral self-adaptive absorber/emitter for harvesting energy from the sun and outer space

The sun (~6000 K) and outer space (~3 K) are the original heat source and sink for human beings on Earth. The energy applications of absorbing solar irradiation and harvesting the coldness of outer space for energy utilization have attracted considerable interest from researchers. However, combining these two functions in a static device for continuous energy harvesting is unachievable due to the intrinsic infrared spectral conflict. In this study, we developed spectral self-adaptive absorber/emitter (SSA/E) for daytime photothermal and nighttime radiative sky cooling modes depending on the phase transition of the vanadium dioxide coated layer. A 24-hour day-night test showed that the fabricated SSA/E has continuous energy harvesting ability and improved overall energy utilization performance, thus showing remarkable potential in future energy applications.

cond-mat.mtrl-sci↗
arXiv2020

Automatically growing global reactive neural network potential energy surfaces: a trajectory free active learning strategy

An efficient and trajectory-free active learning method is proposed to automatically sample data points for constructing globally accurate reactive potential energy surfaces (PESs) using neural networks (NNs). Although NNs do not provide the predictive variance as the Gaussian process regression does, we can alternatively minimize the negative of the squared difference surface (NSDS) given by two different NN models to actively locate the point where the PES is least confident. A batch of points in the minima of this NSDS can be iteratively added into the training set to improve the PES. The configuration space is gradually and globally covered with no need to run classical trajectory (or equivalently molecular dynamics) simulations. Through refitting the available analytical PESs of H3 and OH3 reactive systems, we demonstrate the efficiency and robustness of this new strategy, which enables fast convergence of the reactive PESs with respect to the number of points in terms of quantum scattering probabilities.

physics.chem-ph↗
arXiv2019

Travel Time Estimation without Road Networks: An Urban Morphological Layout Representation Approach

Travel time estimation is a crucial task for not only personal travel scheduling but also city planning. Previous methods focus on modeling toward road segments or sub-paths, then summing up for a final prediction, which have been recently replaced by deep neural models with end-to-end training. Usually, these methods are based on explicit feature representations, including spatio-temporal features, traffic states, etc. Here, we argue that the local traffic condition is closely tied up with the land-use and built environment, i.e., metro stations, arterial roads, intersections, commercial area, residential area, and etc, yet the relation is time-varying and too complicated to model explicitly and efficiently. Thus, this paper proposes an end-to-end multi-task deep neural model, named Deep Image to Time (DeepI2T), to learn the travel time mainly from the built environment images, a.k.a. the morphological layout images, and showoff the new state-of-the-art performance on real-world datasets in two cities. Moreover, our model is designed to tackle both path-aware and path-blind scenarios in the testing phase. This work opens up new opportunities of using the publicly available morphological layout images as considerable information in multiple geography-related smart city applications.

cs.AI↗
arXiv2019

C^3 Framework: An Open-source PyTorch Code for Crowd Counting

This technical report attempts to provide efficient and solid kits addressed on the field of crowd counting, which is denoted as Crowd Counting Code Framework (C$^3$F). The contributions of C$^3$F are in three folds: 1) Some solid baseline networks are presented, which have achieved the state-of-the-arts. 2) Some flexible parameter setting strategies are provided to further promote the performance. 3) A powerful log system is developed to record the experiment process, which can enhance the reproducibility of each experiment. Our code is made publicly available at \url{https://github.com/gjy3035/C-3-Framework}. Furthermore, we also post a Chinese blog\footnote{\url{https://zhuanlan.zhihu.com/p/65650998}} to describe the details and insights of crowd counting.

cs.CV↗
arXiv2019

Mapping System Level Behaviors with Android APIs via System Call Dependence Graphs

Due to Android's open source feature and low barriers to entry for developers, millions of developers and third-party organizations have been attracted into the Android ecosystem. However, over 90 percent of mobile malware are found targeted on Android. Though Android provides multiple security features and layers to protect user data and system resources, there are still some over-privileged applications in Google Play Store or third-party Android app stores at wild. In this paper, we proposed an approach to map system level behavior and Android APIs, based on the observation that system level behaviors cannot be avoided but sensitive Android APIs could be evaded. To the best of our knowledge, our approach provides the first work to map system level behavior and Android APIs through System Call Dependence Graphs. The study also shows that our approach can effectively identify potential permission abusing, with almost negligible performance impact.

cs.CR↗
arXiv2019

Hierarchical Recurrent Neural Network for Video Summarization

Exploiting the temporal dependency among video frames or subshots is very important for the task of video summarization. Practically, RNN is good at temporal dependency modeling, and has achieved overwhelming performance in many video-based tasks, such as video captioning and classification. However, RNN is not capable enough to handle the video summarization task, since traditional RNNs, including LSTM, can only deal with short videos, while the videos in the summarization task are usually in longer duration. To address this problem, we propose a hierarchical recurrent neural network for video summarization, called H-RNN in this paper. Specifically, it has two layers, where the first layer is utilized to encode short video subshots cut from the original video, and the final hidden state of each subshot is input to the second layer for calculating its confidence to be a key subshot. Compared to traditional RNNs, H-RNN is more suitable to video summarization, since it can exploit long temporal dependency among frames, meanwhile, the computation operations are significantly lessened. The results on two popular datasets, including the Combined dataset and VTW dataset, have demonstrated that the proposed H-RNN outperforms the state-of-the-arts.

cs.CV↗
arXiv2019

A General Framework for Edited Video and Raw Video Summarization

In this paper, we build a general summarization framework for both of edited video and raw video summarization. Overall, our work can be divided into three folds: 1) Four models are designed to capture the properties of video summaries, i.e., containing important people and objects (importance), representative to the video content (representativeness), no similar key-shots (diversity) and smoothness of the storyline (storyness). Specifically, these models are applicable to both edited videos and raw videos. 2) A comprehensive score function is built with the weighted combination of the aforementioned four models. Note that the weights of the four models in the score function, denoted as property-weight, are learned in a supervised manner. Besides, the property-weights are learned for edited videos and raw videos, respectively. 3) The training set is constructed with both edited videos and raw videos in order to make up the lack of training data. Particularly, each training video is equipped with a pair of mixing-coefficients which can reduce the structure mess in the training set caused by the rough mixture. We test our framework on three datasets, including edited videos, short raw videos and long raw videos. Experimental results have verified the effectiveness of the proposed framework.

cs.CV↗
arXiv2019

A CNN-RNN Architecture for Multi-Label Weather Recognition

Weather Recognition plays an important role in our daily lives and many computer vision applications. However, recognizing the weather conditions from a single image remains challenging and has not been studied thoroughly. Generally, most previous works treat weather recognition as a single-label classification task, namely, determining whether an image belongs to a specific weather class or not. This treatment is not always appropriate, since more than one weather conditions may appear simultaneously in a single image. To address this problem, we make the first attempt to view weather recognition as a multi-label classification task, i.e., assigning an image more than one labels according to the displayed weather conditions. Specifically, a CNN-RNN based multi-label classification approach is proposed in this paper. The convolutional neural network (CNN) is extended with a channel-wise attention model to extract the most correlated visual features. The Recurrent Neural Network (RNN) further processes the features and excavates the dependencies among weather classes. Finally, the weather labels are predicted step by step. Besides, we construct two datasets for the weather recognition task and explore the relationships among different weather conditions. Experimental results demonstrate the superiority and effectiveness of the proposed approach. The new constructed datasets will be available at https://github.com/wzgwzg/Multi-Label-Weather-Recognition.

cs.CV↗
arXiv2019

Dataflow-based Joint Quantization of Weights and Activations for Deep Neural Networks

This paper addresses a challenging problem - how to reduce energy consumption without incurring performance drop when deploying deep neural networks (DNNs) at the inference stage. In order to alleviate the computation and storage burdens, we propose a novel dataflow-based joint quantization approach with the hypothesis that a fewer number of quantization operations would incur less information loss and thus improve the final performance. It first introduces a quantization scheme with efficient bit-shifting and rounding operations to represent network parameters and activations in low precision. Then it restructures the network architectures to form unified modules for optimization on the quantized model. Extensive experiments on ImageNet and KITTI validate the effectiveness of our model, demonstrating that state-of-the-art results for various tasks can be achieved by this quantized model. Besides, we designed and synthesized an RTL model to measure the hardware costs among various quantization methods. For each quantization operation, it reduces area cost by about 15 times and energy consumption by about 9 times, compared to a strong baseline.

cs.LG↗
arXiv2017

Unique catalytic decomposition and reduction of hydrogen peroxide and organic peroxides on magnesia (001) promoted by oxide-metal hybrid nanostructure

The detection, removal and reduction of hydrogen peroxide is of significant importance for its increasing application in the areas of environment, food, electrochemistry and clinical laboratory. Herein the dissociative adsorption behavior of hydrogen peroxide on ultrathin magnesia (001) films deposited on transition metal is uncovered for the first time by employing periodic density-functional theory calculations with van der Waals corrections. The hydrogen peroxide is dissociated smoothly and reduced to surface hydroxyls on MgO(001)/TM, and the dissociative adsorption energies of all the considered fragmentation configurations are substantially negative, demonstrating dissociation and reduction of hydrogen peroxide on MgO(001)/TM is thermodynamically favorable. The dissociative adsorption energy decreases monotonously with increasing film thickness, demonstrating the lower reactivity of thick oxide films. Moreover, we examined the suitability of several transition metal slabs (molybdenum, silver, vanadium, tungsten and gold) combined with magnesia for splitting hydrogen peroxide. The dissociative adsorption energies enhance when the lattice constants of substrate slabs increase, indicating the chemisorption strength can be tuned by category of metal slabs as well as thickness of oxide film. The mechanism of reactivity enhancement for energetically and dynamically favorable decomposition of hydrogen peroxide on supported magnesia is elucidated by characterizing the geometric structures and electronic properties. The fragmentation and reduction of diethyl peroxide and peroxyacetone are also studied to reveal the catalytic activity of ultrathin magnesia toward splitting organic peroxides. The results are wished to provide useful clue for detecting and reducing hydrogen peroxide and organic peroxides by employing oxide-metal hybrid nanostructure.

physics.chem-ph↗
arXiv2017

Controllability of Heterogeneous Multi-Agent Networks

The existing results on controllability of multi-agents networks are mostly based on homogeneous nodes. This paper focuses on controllability of heterogeneous multi-agent networks, where the agents are modeled as two types. One type is that the agents are of the same high-order dynamics, and the interconnection topologies of the information flow in different orders are supposed to be different. It is proved that a heterogeneous-topology network is controllable if and only if the first-order information topology is leader-follower connected, and there exists a Laplacian matrix, which is a linear combination of the Laplacian matrices of each order information, whose corresponding topology is controllable. The other type is that the agents are of generic linear dynamics, and the dynamics are supposed to be heterogeneous. A necessary and sufficient condition for controllability of heterogeneous-dynamic networks is that each agent contains a controllable dynamic part, and the interconnection topology of the network is leader-follower connected. If some dynamics of the agents are not controllable, the controllability between the agents and the whole network is also studied by introducing the concept of eigenvector-uncontrollable. Different illustrative examples are provided to demonstrate the effectiveness of the theoretical results in this paper.

math.OC↗
arXiv2017

Remarkably strong chemisorption of nitric oxide on insulating oxide films promoted by hybrid structure

The remarkably strong chemical adsorption behaviors of nitric oxide on magnesia (001) film deposited on metal substrate have been investigated by employing periodic density functional calculations with Van der Waals corrections. The molybdenum supported magnesia (001) show significantly enhanced adsorption properties and the nitric oxide is chemisorbed strongly and preferably trapped in flat adsorption configuration on metal supported oxide film, due to the substantially large adsorption energies and transformation barriers. The analysis of Bader charges, projected density of states, differential charge densities, electron localization function, highest occupied orbital and particular orbital with largest Mg-NO-Mg bonding coefficients, are applied to reveal the electronic adsorption properties and characteristics of bonding between nitric oxide and surface as well as the bonding within the hybrid structure. The strong chemical binding of nitric oxide on magnesia deposited on molybdenum slab offers new opportunities for toxic gas detection and treatment. We anticipate that hybrid structure promoted remarkable chemical adsorption of nitric oxide on magnesia in this study will provide versatile strategy for enhancing chemical reactivity and properties of insulating oxide.

physics.chem-ph↗
arXiv2016

Non-Fragility and Partial Controllability of Multi-Agent Systems

Controllability of multi-agent systems is determined by the interconnection topologies. In practice, losing agents can change the topologies of multi-agent systems, which may affect the controllability. This paper studies non-fragility of controllability influenced by losing agents. In virtue of the concept of cutsets, necessary and sufficient conditions are established from a graphic perspective, for strong non-fragility and weak non-fragility of controllability, respectively. For multi-agent systems which contain important agents, partial controllability is proposed in terms of the concept of controllable node groups, and necessary and sufficient criteria are established for partial controllability. Moreover, partial controllability preserving problem is proposed. Utilizing the concept of compressed graphs, this problem is transformed into finding the the minimal $\mathbf{\langle s,t\rangle}$ vertex cutsets of the interconnection graph, which has a polynomial-time complexity algorithm for the solution. Several constructive examples illuminate the theoretical results.

eess.SY↗
arXiv2015

Leader selection and weight adjustment problems for multi-agent systems

For an uncontrollable system, adding leaders and adjusting edge weights are two methods to improve controllability. In this paper, controllability of multi-agent systems under directed topologies is studied, especially on leader selection problem and weight adjustment problem. For a given system, necessary and sufficient algebraic conditions for controllability with fewest leaders are proposed. From another perspective, when leaders are fixed, controllability could be improved by adjusting edge weights, and therefore the system is supposed to be structurally controllable, which holds if and only if the communication topology contains a spanning tree. It is also proved that the number of fewest edges needed to be assigned on new weights equals the rank deficiency of controllability matrix. An algorithm on how to perform weight adjustment is presented. Simulation examples are provided to illustrate the theoretical results.

eess.SY↗