Showing papers by "Tongji University published in 2014"

Saliency Optimization from Robust Background Detection

Abstract: Recent progresses in salient object detection have exploited the boundary prior, or background information, to assist other saliency cues such as contrast, achieving state-of-the-art results. However, their usage of boundary prior is very simple, fragile, and the integration with other cues is mostly heuristic. In this work, we present new methods to address these issues. First, we propose a robust background measure, called boundary connectivity. It characterizes the spatial layout of image regions with respect to image boundaries and is much more robust. It has an intuitive geometrical interpretation and presents unique benefits that are absent in previous saliency measures. Second, we propose a principled optimization framework to integrate multiple low level cues, including our background measure, to obtain clean and uniform saliency maps. Our formulation is intuitive, efficient and achieves state-of-the-art results on several benchmark datasets.

Inflammatory bowel disease: pathogenesis.

Abstract: Inflammatory bowel disease (IBD), including Crohn's disease and ulcerative colitis, is characterized by chronic relapsing intestinal inflammation It has been a worldwide health-care problem with a continually increasing incidence It is thought that IBD results from an aberrant and continuing immune response to the microbes in the gut, catalyzed by the genetic susceptibility of the individual Although the etiology of IBD remains largely unknown, it involves a complex interaction between the genetic, environmental or microbial factors and the immune responses Of the four components of IBD pathogenesis, most rapid progress has been made in the genetic study of gut inflammation The latest internationally collaborative studies have ascertained 163 susceptibility gene loci for IBD The genes implicated in childhood-onset and adult-onset IBD overlap, suggesting similar genetic predispositions However, the fact that genetic factors account for only a portion of overall disease variance indicates that microbial and environmental factors may interact with genetic elements in the pathogenesis of IBD Meanwhile, the adaptive immune response has been classically considered to play a major role in the pathogenesis of IBD, as new studies in immunology and genetics have clarified that the innate immune response maintains the same importance in inducing gut inflammation Recent progress in understanding IBD pathogenesis sheds lights on relevant disease mechanisms, including the innate and adaptive immunity, and the interactions between genetic factors and microbial and environmental cues In this review, we provide an update on the major advances that have occurred in above areas

VSI: a visual saliency-induced index for perceptual image quality assessment.

Abstract: Perceptual image quality assessment (IQA) aims to use computational models to measure the image quality in consistent with subjective evaluations. Visual saliency (VS) has been widely studied by psychologists, neurobiologists, and computer scientists during the last decade to investigate, which areas of an image will attract the most attention of the human visual system. Intuitively, VS is closely related to IQA in that suprathreshold distortions can largely affect VS maps of images. With this consideration, we propose a simple but very effective full reference IQA method using VS. In our proposed IQA model, the role of VS is twofold. First, VS is used as a feature when computing the local quality map of the distorted image. Second, when pooling the quality score, VS is employed as a weighting function to reflect the importance of a local region. The proposed IQA index is called visual saliency-based index (VSI). Several prominent computational VS models have been investigated in the context of IQA and the best one is chosen for VSI. Extensive experiments performed on four large-scale benchmark databases demonstrate that the proposed IQA index VSI works better in terms of the prediction accuracy than all state-of-the-art IQA indices we can find while maintaining a moderate computational complexity. The MATLAB source code of VSI and the evaluation results are publicly available online at http://sse.tongji.edu.cn/linzhang/IQA/VSI/VSI.htm.

Functional surface engineering of C-dots for fluorescent biosensing and in vivo bioimaging.

Abstract: Nanoparticles are promising scaffolds for applications such as imaging, chemical sensors and biosensors, diagnostics, drug delivery, catalysis, energy, photonics, medicine, and more. Surface functionalization of nanoparticles introduces an additional dimension in controlling nanoparticle interfacial properties and provides an effective bridge to connect nanoparticles to biological systems. With fascinating photoluminescence properties, carbon dots (C-dots), carbon-containing nanoparticles that are attracting considerable attention as a new type of quantum dot, are becoming both an important class of imaging probes and a versatile platform for engineering multifunctional nanosensors. In order to transfer C-dots from proof-of-concept studies toward real world applications such as in vivo bioimaging and biosensing, careful design and engineering of C-dot probes is becoming increasingly important.A comprehensive knowledge of how C-dot surfaces with various properties behave is essential for engineering C-dots...

Metal Nanowire Networks: The Next Generation of Transparent Conductors

Abstract: There is an ongoing drive to replace the most common transparent conductor, indium tin oxide (ITO), with a material that gives comparable performance, but can be coated from solution at speeds orders of magnitude faster than the sputtering processes used to deposit ITO. Metal nanowires are currently the only alternative to ITO that meets these requirements. This Progress Report summarizes recent advances toward understanding the relationship between the structure of metal nanowires, the electrical and optical properties of metal nanowires, and the properties of a network of metal nanowires. Using the structure–property relationship of metal nanowire networks as a roadmap, this Progress Report describes different synthetic strategies to produce metal nanowires with the desired properties. Practical aspects of processing metal nanowires into high-performance transparent conducting films are discussed, as well as the use of nanowire films in a variety of applications.

XBP1 promotes triple-negative breast cancer by controlling the HIF1α pathway

Abstract: Cancer cells induce a set of adaptive response pathways to survive in the face of stressors due to inadequate vascularization. One such adaptive pathway is the unfolded protein (UPR) or endoplasmic reticulum (ER) stress response mediated in part by the ER-localized transmembrane sensor IRE1 (ref. 2) and its substrate XBP1 (ref. 3). Previous studies report UPR activation in various human tumours, but the role of XBP1 in cancer progression in mammary epithelial cells is largely unknown. Triple-negative breast cancer (TNBC)--a form of breast cancer in which tumour cells do not express the genes for oestrogen receptor, progesterone receptor and HER2 (also called ERBB2 or NEU)--is a highly aggressive malignancy with limited treatment options. Here we report that XBP1 is activated in TNBC and has a pivotal role in the tumorigenicity and progression of this human breast cancer subtype. In breast cancer cell line models, depletion of XBP1 inhibited tumour growth and tumour relapse and reduced the CD44(high)CD24(low) population. Hypoxia-inducing factor 1α (HIF1α) is known to be hyperactivated in TNBCs. Genome-wide mapping of the XBP1 transcriptional regulatory network revealed that XBP1 drives TNBC tumorigenicity by assembling a transcriptional complex with HIF1α that regulates the expression of HIF1α targets via the recruitment of RNA polymerase II. Analysis of independent cohorts of patients with TNBC revealed a specific XBP1 gene expression signature that was highly correlated with HIF1α and hypoxia-driven signatures and that strongly associated with poor prognosis. Our findings reveal a key function for the XBP1 branch of the UPR in TNBC and indicate that targeting this pathway may offer alternative treatment strategies for this aggressive subtype of breast cancer.

Observer-Based Output Feedback Event-Triggered Control for Consensus of Multi-Agent Systems

Abstract: This paper studies the consensus problem of linear multi-agent systems via observer-based event-triggered control. Two novel observer-based event-triggered control schemes, one centralized and the other distributed, are developed. It is shown that under the proposed control protocols, consensus can be reached if the underlying communication graph of the MAS is connected. An example is finally presented to illustrate the effectiveness of the proposed control methods.

Experimental demonstration of a bilayer thermal cloak

Abstract: Invisibility has attracted intensive research in various communities, e.g., optics, electromagnetics, acoustics, thermodynamics, dc, etc. However, many experimental demonstrations have only been achieved by virtue of simplified approaches due to the inhomogeneous and extreme parameters imposed by the transformation-optic method, and usually require a challenging realization with metamaterials. In this Letter, we demonstrate a bilayer thermal cloak made of bulk isotropic materials, and it has been validated as an exact cloak. We experimentally verified its ability to maintain the heat front and its heat protection capabilities in a 2D proof-of-concept experiment. The robustness of this scheme is validated in both 2D (including oblique heat front incidence) and 3D configurations. The proposed scheme may open a new avenue to control the diffusive heat flow in ways inconceivable with phonons, and also inspire new alternatives to the functionalities promised by transformation optics.

Experimental Demonstration of a Coherent Perfect Absorber with PT Phase Transition

Abstract: We report the realization of a coherent perfect absorber, using a pair of passive resonators coupled to a microwave transmission line in the background, which can completely absorb light in its parity-time (PT-)symmetric phase but not in its broken phase. Instead of balancing material gain and loss, we exploit the incident waves in the open system as an effective gain so that ideal PT symmetry can be established by using only passive materials. Such a route will be effective to construct PT-symmetric metamaterials and also tunable PT-symmetric optical elements in general. It also provides a flexible platform for studying exceptional-point physics with both electric and magnetic responses.

Robust and Efficient Subspace Segmentation via Least Squares Regression

Abstract: This paper studies the subspace segmentation problem which aims to segment data drawn from a union of multiple linear subspaces. Recent works by using sparse representation, low rank representation and their extensions attract much attention. If the subspaces from which the data drawn are independent or orthogonal, they are able to obtain a block diagonal affinity matrix, which usually leads to a correct segmentation. The main differences among them are their objective functions. We theoretically show that if the objective function satisfies some conditions, and the data are sufficiently drawn from independent subspaces, the obtained affinity matrix is always block diagonal. Furthermore, the data sampling can be insufficient if the subspaces are orthogonal. Some existing methods are all special cases. Then we present the Least Squares Regression (LSR) method for subspace segmentation. It takes advantage of data correlation, which is common in real data. LSR encourages a grouping effect which tends to group highly correlated data together. Experimental results on the Hopkins 155 database and Extended Yale Database B show that our method significantly outperforms state-of-the-art methods. Beyond segmentation accuracy, all experiments demonstrate that LSR is much more efficient.

Ages and magnetic structures of the South China Sea constrained by deep tow magnetic surveys and IODP Expedition 349

Abstract: Combined analyses of deep tow magnetic anomalies and International Ocean Discovery Program Expedition 349 cores show that initial seafloor spreading started around 33 Ma in the northeastern South China Sea (SCS), but varied slightly by 1-2 Myr along the northern continent-ocean boundary (COB). A southward ridge jump of approximate to 20 km occurred around 23.6 Ma in the East Subbasin; this timing also slightly varied along the ridge and was coeval to the onset of seafloor spreading in the Southwest Subbasin, which propagated for about 400 km southwestward from approximate to 23.6 to approximate to 21.5 Ma. The terminal age of seafloor spreading is approximate to 15 Ma in the East Subbasin and approximate to 16 Ma in the Southwest Subbasin. The full spreading rate in the East Subbasin varied largely from approximate to 20 to approximate to 80 km/Myr, but mostly decreased with time except for the period between approximate to 26.0 Ma and the ridge jump (approximate to 23.6 Ma), within which the rate was the fastest at approximate to 70 km/Myr on average. The spreading rates are not correlated, in most cases, to magnetic anomaly amplitudes that reflect basement magnetization contrasts. Shipboard magnetic measurements reveal at least one magnetic reversal in the top 100 m of basaltic layers, in addition to large vertical intensity variations. These complexities are caused by late-stage lava flows that are magnetized in a different polarity from the primary basaltic layer emplaced during the main phase of crustal accretion. Deep tow magnetic modeling also reveals this smearing in basement magnetizations by incorporating a contamination coefficient of 0.5, which partly alleviates the problem of assuming a magnetic blocking model of constant thickness and uniform magnetization. The primary contribution to magnetic anomalies of the SCS is not in the top 100 m of the igneous basement.

Ni12P5 Nanoparticles as an Efficient Catalyst for Hydrogen Generation via Electrolysis and Photoelectrolysis

Abstract: The exploitation of a low-cost catalyst is desirable for hydrogen generation from electrolysis or photoelectrolysis. In this study we have demonstrated that nickel phosphide (Ni12P5) nanoparticles have efficient and stable catalytic activity for the hydrogen evolution reaction. The catalytic performance of Ni12P5 nanoparticles is favorably comparable to those of recently reported efficient nonprecious catalysts. The optimal overpotential required for 20 mA/cm(2) current density is 143 ± 3 mV in acidic solution (H2SO4, 0.5 M). The catalytic activity of Ni12P5 is likely to be correlated with the charged natures of Ni and P. Ni12P5 nanoparticles were introduced to silicon nanowires, and the power conversion efficiency of the resulting composite is larger than that of silicon nanowires decorated with platinum particles. This result demonstrates the promising application potential of metal phosphide in photoelectrochemical hydrogen generation.

Phonon-glass electron-crystal thermoelectric clathrates : Experiments and theory

Abstract: Type-I clathrate compounds have attracted a great deal of interest in connection with the search for efficient thermoelectric materials. These compounds constitute networked cages consisting of nano-scale tetrakaidecahedrons (14 hedrons) and dodecahedrons (12 hedrons), in which the group 1 or 2 elements in the periodic table are encaged as the so-called rattling guest atom. It is remarkable that, though these compounds have crystalline cubic-structure, they exhibit glass-like phonon thermal conductivity over the whole temperature range depending on the states of rattling guest atoms in the tetrakaidecahedron. In addition, these compounds show unusual glass-like specific heats and THz-frequency phonon dynamics, providing a remarkable broad peak almost identical to those observed in topologically disordered amorphous materials or structural glasses, the so-called Boson peak. An efficient thermoelectric effect is realized in compounds showing these glass-like characteristics. This decade, a number of experimental works dealing with type-I clathrate compounds have been published. These are diffraction experiments, thermal and spectroscopic experiments in addition to those based on heat and electronic transport. These form the raw materials for this article based on advances this decade. The subject of this article involves interesting phenomena from the viewpoint of not only physics but also from the view point of the practical problem of elaborating efficient thermoelectric materials. This review presents a survey of a wide range of experimental investigations of type-I clathrate compounds, together with a review of theoretical interpretations of the peculiar thermal and dynamic properties observed in these materials.

Three-Dimensional Co3O4@MnO2 Hierarchical Nanoneedle Arrays: Morphology Control and Electrochemical Energy Storage

Abstract: In this paper, a highly ordered three-dimensional Co3O4@MnO2 hierarchical porous nanoneedle array on nickel foam is fabricated by a facile, stepwise hydrothermal approach. The morphologies evolution of Co3O4 and Co3O4@MnO2 nanostructures upon reaction times and growth temperature are investigated in detail. Moreover, the as-prepared Co3O4@MnO2 hierarchical structures are investigated as anodes for both supercapacitors and Li-ion batteries. When used for supercapacitors, excellent electrochemical performances such as high specific capacitances of 932.8 F g−1 at a scan rate of 10 mV s−1 and 1693.2 F g−1 at a current density of 1 A g−1 as well as long-term cycling stability and high energy density (66.2 W h kg−1 at a power density of 0.25 kW kg−1), which are better than that of the individual component of Co3O4 nanoneedles and MnO2 nanosheets, are obtained. The Co3O4@MnO2 NAs are also tested as anode material for LIBs for the first time, which presents an improved performance with high reversible capacity of 1060 mA h g−1 at a rate of 120 mA g−1, good cycling stability, and rate capability.

Comparative analysis of metazoan chromatin organization

Abstract: Genome function is dynamically regulated in part by chromatin, which consists of the histones, non-histone proteins and RNA molecules that package DNA. Studies in Caenorhabditis elegans and Drosophila melanogaster have contributed substantially to our understanding of molecular mechanisms of genome function in humans, and have revealed conservation of chromatin components and mechanisms. Nevertheless, the three organisms have markedly different genome sizes, chromosome architecture and gene organization. On human and fly chromosomes, for example, pericentric heterochromatin flanks single centromeres, whereas worm chromosomes have dispersed heterochromatin-like regions enriched in the distal chromosomal 'arms', and centromeres distributed along their lengths. To systematically investigate chromatin organization and associated gene regulation across species, we generated and analysed a large collection of genome-wide chromatin data sets from cell lines and developmental stages in worm, fly and human. Here we present over 800 new data sets from our ENCODE and modENCODE consortia, bringing the total to over 1,400. Comparison of combinatorial patterns of histone modifications, nuclear lamina-associated domains, organization of large-scale topological domains, chromatin environment at promoters and enhancers, nucleosome positioning, and DNA replication patterns reveals many conserved features of chromatin organization among the three organisms. We also find notable differences in the composition and locations of repressive chromatin. These data sets and analyses provide a rich resource for comparative and species-specific investigations of chromatin composition, organization and function.

Magnetic ordered mesoporous copper ferrite as a heterogeneous Fenton catalyst for the degradation of imidacloprid

Abstract: Highly ordered mesoporous copper ferrite (meso-CuFe2O4) with high surface area and large pore size was successfully fabricated and firstly proposed as a heterogeneous Fenton catalyst. It was synthesized through the nanocasting strategy by using KIT-6 as hard template. The morphology and physicochemical properties of meso-CuFe2O4 were characterized by SEM, TEM, XRD, XPS, FT-IR, Raman spectra, etc. The obtained results showed that the surface area and the pore size of meso-CuFe2O4 were 122 m2 g−1 and 9.2 nm, respectively. The meso-CuFe2O4 presented excellent catalytic activity for the degradation of imidacloprid, achieving almost complete removal of 10 mg L−1 imidacloprid after 5 h at the reaction conditions of 0.3 g L−1 catalyst and 40 mM H2O2. Kinetic analysis showed that the degradation of imidacloprid follows the pseudo-first order. The apparent rate constant for meso-CuFe2O4 was 1.0445 h−1, which was 1.5, 2 and 2.5 times than those of meso-CoFe2O4, con-CuFe2O4 and nano-Fe3O4, respectively. The amount of hydroxyl radical (OH) generated was directly proportional to the degradation efficiency of imidacloprid, suggesting the involvement of OH in oxidizing imidacloprid. The obtained results indicated that meso-CuFe2O4 presented the highest activity, which was not only due to its ordered mesoporous structure with high surface area and large pore size, but also assigned to the redox recycle of Fe2+/Fe3+ and Cu+/Cu2+ in meso-CuFe2O4. The special effect of Cu was discussed in terms of the thermodynamically favorable Fe3+ reduction by Cu+, regenerating the active species Fe2+. The meso-CuFe2O4 presented very low iron leaching (<1 ppm) even in acidic condition and retained almost its high catalytic activity after 5 consecutive runs. Besides, meso-CuFe2O4 possessed medium saturation magnetization, which had provided a potential advantage for the recovery and reuse of catalyst.

The WASOG Sarcoidosis Organ Assessment Instrument: An update of a previous clinical tool

Abstract: Introduction: A Case Control Etiology of Sarcoidosis Study (ACCESS) sarcoidosis organ assessment instrument has been used for more than a decade to establish uniform standards for the probability of sarcoidosis organ involvement. The ACCESS instrument has become increasingly outdated as new technologies have been developed. Furthermore, the ACCESS instrument failed to address all possible organs involved with sarcoidosis. For these reasons, the World Association of Sarcoidosis and Other Granulomatous Diseases (WASOG) developed a new sarcoidosis organ assessment instrument. Methods: Clinical sarcoidosis experts assessed various clinical manifestations for the probability of sarcoidosis organ involvement. Two criteria were required to apply this assessment: 1) histologic evidence of granulomatous inflammation of unknown cause in an organ that was not being assessed; 2) the clinical manifestation being addressed required that alternative causes other than sarcoidosis had been reasonably excluded. Clinical manifestations were assessed as either: a) highly probable: likelihood of sarcoidosis causing this manifestation of at least 90%.; b) probable: likelihood of sarcoidosis causing this manifestation of between 50 and 90%; c) possible: likelihood of sarcoidosis causing this manifestation of less than 50%. The sarcoidosis experts voted on the likelihood of sarcoidosis causing each manifestation using Delphi study methodology where at least 70% agreement of the experts was needed for consensus. Results: Various clinical manifestations were classified as highly probable, at least probable, possible, or indeterminate when no consensus could be reached. Conclusion: An instrument was developed by expert opinion that may be useful for the clinician and researcher in establishing criteria for sarcoidosis organ involvement.

Quality assessment for building footprints data on OpenStreetMap

Abstract: In the past two years, several applications of generating three-dimensional 3D buildings from OpenStreetMap OSM have been made available, for instance, OSM-3D, OSM2World, OSM Building, etc In these projects, 3D buildings are reconstructed using the buildings’ footprints and information about their attributes, which are documented as tags in OSM Therefore, the quality of 3D buildings relies strongly on the quality of the building footprints data in OSM This article is dedicated to a quality assessment of building footprints data in OSM for the German city of Munich, which is one of the most developed cities in OSM The data are evaluated in terms of completeness, semantic accuracy, position accuracy, and shape accuracy by using building footprints in ATKIS German Authority Topographic–Cartographic Information System as reference data The process contains three steps: finding correspondence between OSM and ATKIS data, calculating parameters of the four quality criteria, and statistical analysis The results show that OSM footprint data in Munich have a high completeness and semantic accuracy There is an offset of about four meters on average in terms of position accuracy With respect to shape, OSM building footprints have a high similarity to those in ATKIS data However, some architectural details are missing; hence, the OSM footprints can be regarded as a simplified version of those in ATKIS data

Optimum Carrier Concentration in n-Type PbTe Thermoelectrics

Abstract: Taking La- and I-doped PbTe as an example, the current work shows the effects of optimizing the thermoelectric figure of merit, zT, by controlling the doping level. The high doping effectiveness allows the carrier concentration to be precisely designed and prepared to control the Fermi level. In addition to the Fermi energy tuning, La-doping modifies the conduction band, leading to an increase in the density of states effective mass that is confirmed by transport, infrared reflectance and hard X-ray photoelectron spectroscopy measurements. Taking such a band structure modification effect into account, the electrical transport properties can then be well-described by a self-consistent single non-parabolic Kane band model that yields an approximate (m*T)^(1.5) dependence of the optimal carrier concentration for a peak power factor in both doping cases. Such a simple temperature dependence also provides an effective approximation of carrier concentration for a peak zT and helps to explain, the effects of other strategies such as lowering the lattice thermal conductivity by nanostructuring or alloying in n-PbTe, which demonstrates a practical guide for fully optimizing thermoelectric materials in the entire temperature range. The principles used here should be equally applicable to other thermoelectric materials.

Context-aware vehicular cyber-physical systems with cloud support: architecture, challenges, and solutions

Abstract: The advances in wireless communication techniques, mobile cloud computing, and context- aware technologies boost a growing interest in the design, development, and deployment of vehicular networks for emerging applications. This leads to an increasing evolutionary tendency to change from vehicular networks toward cloud-assisted context-aware vehicular cyberphysical systems. In this article, we first propose a multi-layered context-aware architecture and introduce two crucial service components, vehicular social networks and context-aware vehicular security. Then we propose an application scenario regarding the context-aware dynamic parking services by illuminating the cloud-assisted architecture and logic flow. Finally, we investigate the challenges and possible solutions, including context-aware safety hazard prediction, context-aware dynamic vehicle routing, and context-aware vehicular clouds.

Cleavage of tau by asparagine endopeptidase mediates the neurofibrillary pathology in Alzheimer’s disease

Abstract: Neurofibrillary tangles (NFTs), composed of truncated and hyperphosphorylated tau, are a common feature of numerous aging-related neurodegenerative diseases, including Alzheimer's disease (AD). However, the molecular mechanisms mediating tau truncation and aggregation during aging remain elusive. Here we show that asparagine endopeptidase (AEP), a lysosomal cysteine proteinase, is activated during aging and proteolytically degrades tau, abolishes its microtubule assembly function, induces tau aggregation and triggers neurodegeneration. AEP is upregulated and active during aging and is activated in human AD brain and tau P301S-transgenic mice with synaptic pathology and behavioral impairments, leading to tau truncation in NFTs. Tau P301S-transgenic mice with deletion of the gene encoding AEP show substantially reduced tau hyperphosphorylation, less synapse loss and rescue of impaired hippocampal synaptic function and cognitive deficits. Mice infected with adeno-associated virus encoding an uncleavable tau mutant showed attenuated pathological and behavioral defects compared to mice injected with adeno-associated virus encoding tau P301S. Together, these observations indicate that AEP acts as a crucial mediator of tau-related clinical and neuropathological changes. Inhibition of AEP may be therapeutically useful for treating tau-mediated neurodegenerative diseases.

Full control and manipulation of heat signatures: cloaking, camouflage and thermal metamaterials.

Abstract: Thermal camouflage and cloaking can transform an actual heat signature into a pre-controlled one. A viable recipe for controlling and manipulating heat signatures using thermal metamaterials to empower cloaking and camouflage in heat conduction is demonstrated. The thermal signature of the object is thus metamorphosed and perceived as multiple targets with different geometries and compositions, with the original object cloaked.

Mixed Pooling for Convolutional Neural Networks

Abstract: Convolutional Neural Network (CNN) is a biologically inspired trainable architecture that can learn invariant features for a number of applications. In general, CNNs consist of alternating convolutional layers, non-linearity layers and feature pooling layers. In this work, a novel feature pooling method, named as mixed pooling, is proposed to regularize CNNs, which replaces the deterministic pooling operations with a stochastic procedure by randomly using the conventional max pooling and average pooling methods. The advantage of the proposed mixed pooling method lies in its wonderful ability to address the over-fitting problem encountered by CNN generation. Experimental results on three benchmark image classification datasets demonstrate that the proposed mixed pooling method is superior to max pooling, average pooling and some other state-of-the-art works known in the literature.