arXiv subjects

Min Xu

Publications and source records attributed to Min Xu.

At least 235 records · Page 13Linked to original sources

arXiv2018

Learning Multi-level Deep Representations for Image Emotion Classification

In this paper, we propose a new deep network that learns multi-level deep representations for image emotion classification (MldrNet). Image emotion can be recognized through image semantics, image aesthetics and low-level visual features from both global and local views. Existing image emotion classification works using hand-crafted features or deep features mainly focus on either low-level visual features or semantic-level image representations without taking all factors into consideration. The proposed MldrNet combines deep representations of different levels, i.e. image semantics, image aesthetics, and low-level visual features to effectively classify the emotion types of different kinds of images, such as abstract paintings and web images. Extensive experiments on both Internet images and abstract paintings demonstrate the proposed method outperforms the state-of-the-art methods using deep features or hand-crafted features. The proposed approach also outperforms the state-of-the-art methods with at least 6% performance improvement in terms of overall classification accuracy.

cs.CV↗

arXiv2018

Adjoint-based optimization for thrust performance of a three-dimensional pitching-rolling plate

An adjoint-based optimization is applied to study the thrust performance of a pitching-rolling ellipsoidal plate in a uniform stream at Reynolds number 100. To achieve the highest thrust, the optimal kinematics of pitching-rolling motion is sought in a large control space including the pitching amplitude, the rolling amplitude, and the phase delay between the pitching and rolling motion. A continuous adjoint approach with boundary motion being handled by non-cylindrical calculous is developed as a computationally efficient optimization algorithm to deal with the large control space with morphing domain. The comparison between the optimal motion and other reference motions shows a significant improvement of thrust from the increase of rolling amplitude and an optimal phase delay of $122.6^\circ$ between the pitching and the rolling motion. The combination of these two factors impacts the overall thrust performance through their strong effects on the angle of attack, circulation, and the pressure distribution on the plate. Further wake structure analysis suggests that the optimal control improves its propulsive performance by generating a stronger leading-edge vortex (LEV) and straightening the wake deflection.

physics.flu-dyn↗

arXiv2018

Dual Pattern Learning Networks by Empirical Dual Prediction Risk Minimization

Motivated by the observation that humans can learn patterns from two given images at one time, we propose a dual pattern learning network architecture in this paper. Unlike conventional networks, the proposed architecture has two input branches and two loss functions. Instead of minimizing the empirical risk of a given dataset, dual pattern learning networks is trained by minimizing the empirical dual prediction loss. We show that this can improve the performance for single image classification. This architecture forces the network to learn discriminative class-specific features by analyzing and comparing two input images. In addition, the dual input structure allows the network to have a considerably large number of image pairs, which can help address the overfitting issue due to limited training data. Moreover, we propose to associate each input branch with a random interest value for learning corresponding image during training. This method can be seen as a stochastic regularization technique, and can further lead to generalization performance improvement. State-of-the-art deep networks can be adapted to dual pattern learning networks without increasing the same number of parameters. Extensive experiments on CIFAR-10, CIFAR- 100, FI-8, Google commands dataset, and MNIST demonstrate that our DPLNets exhibit better performance than original networks. The experimental results on subsets of CIFAR- 10, CIFAR-100, and MNIST demonstrate that dual pattern learning networks have good generalization performance on small datasets.

cs.CV↗

arXiv2018

Respond-CAM: Analyzing Deep Models for 3D Imaging Data by Visualizations

The convolutional neural network (CNN) has become a powerful tool for various biomedical image analysis tasks, but there is a lack of visual explanation for the machinery of CNNs. In this paper, we present a novel algorithm, Respond-weighted Class Activation Mapping (Respond-CAM), for making CNN-based models interpretable by visualizing input regions that are important for predictions, especially for biomedical 3D imaging data inputs. Our method uses the gradients of any target concept (e.g. the score of target class) that flows into a convolutional layer. The weighted feature maps are combined to produce a heatmap that highlights the important regions in the image for predicting the target concept. We prove a preferable sum-to-score property of the Respond-CAM and verify its significant improvement on 3D images from the current state-of-the-art approach. Our tests on Cellular Electron Cryo-Tomography 3D images show that Respond-CAM achieves superior performance on visualizing the CNNs with 3D biomedical images inputs, and is able to get reasonably good results on visualizing the CNNs with natural image inputs. The Respond-CAM is an efficient and reliable approach for visualizing the CNN machinery, and is applicable to a wide variety of CNN model families and image analysis tasks.

cs.CV↗

arXiv2018

Multi-task Learning for Macromolecule Classification, Segmentation and Coarse Structural Recovery in Cryo-Tomography

Cellular Electron Cryo-Tomography (CECT) is a powerful 3D imaging tool for studying the native structure and organization of macromolecules inside single cells. For systematic recognition and recovery of macromolecular structures captured by CECT, methods for several important tasks such as subtomogram classification and semantic segmentation have been developed. However, the recognition and recovery of macromolecular structures are still very difficult due to high molecular structural diversity, crowding molecular environment, and the imaging limitations of CECT. In this paper, we propose a novel multi-task 3D convolutional neural network model for simultaneous classification, segmentation, and coarse structural recovery of macromolecules of interest in subtomograms. In our model, the learned image features of one task are shared and thereby mutually reinforce the learning of other tasks. Evaluated on realistically simulated and experimental CECT data, our multi-task learning model outperformed all single-task learning methods for classification and segmentation. In addition, we demonstrate that our model can generalize to discover, segment and recover novel structures that do not exist in the training data.

cs.CV↗

arXiv2018

Image-derived generative modeling of pseudo-macromolecular structures - towards the statistical assessment of Electron CryoTomography template matching

Cellular Electron CryoTomography (CECT) is a 3D imaging technique that captures information about the structure and spatial organization of macromolecular complexes within single cells, in near-native state and at sub-molecular resolution. Although template matching is often used to locate macromolecules in a CECT image, it is insufficient as it only measures the relative structural similarity. Therefore, it is preferable to assess the statistical credibility of the decision through hypothesis testing, requiring many templates derived from a diverse population of macromolecular structures. Due to the very limited number of known structures, we need a generative model to efficiently and reliably sample pseudo-structures from the complex distribution of macromolecular structures. To address this challenge, we propose a novel image-derived approach for performing hypothesis testing for template matching by constructing generative models using the generative adversarial network. Finally, we conducted hypothesis testing experiments for template matching on both simulated and experimental subtomograms, allowing us to conclude the identity of subtomograms with high statistical credibility and significantly reducing false positives.

q-bio.QM↗

arXiv2018

An integration of fast alignment and maximum-likelihood methods for electron subtomogram averaging and classification

Motivation: Cellular Electron CryoTomography (CECT) is an emerging 3D imaging technique that visualizes subcellular organization of single cells at submolecular resolution and in near-native state. CECT captures large numbers of macromolecular complexes of highly diverse structures and abundances. However, the structural complexity and imaging limits complicate the systematic de novo structural recovery and recognition of these macromolecular complexes. Efficient and accurate reference-free subtomogram averaging and classification represent the most critical tasks for such analysis. Existing subtomogram alignment based methods are prone to the missing wedge effects and low signal-to-noise ratio (SNR). Moreover, existing maximum-likelihood based methods rely on integration operations, which are in principle computationally infeasible for accurate calculation. Results: Built on existing works, we propose an integrated method, Fast Alignment Maximum Likelihood method (FAML), which uses fast subtomogram alignment to sample sub-optimal rigid transformations. The transformations are then used to approximate integrals for maximum-likelihood update of subtomogram averages through expectation-maximization algorithm. Our tests on simulated and experimental subtomograms showed that, compared to our previously developed fast alignment method (FA), FAML is significantly more robust to noise and missing wedge effects with moderate increases of computation cost.Besides, FAML performs well with significantly fewer input subtomograms when the FA method fails. Therefore, FAML can serve as a key component for improved construction of initial structural models from macromolecules captured by CECT.

q-bio.QM↗

arXiv2018

A Telecom Perspective on the Internet of Drones: From LTE-Advanced to 5G

Drones are driving numerous and evolving use cases, and creating transformative socio-economic benefits. Drone operation needs wireless connectivity for communication between drones and ground control systems, among drones, and between drones and air traffic management systems. Mobile networks are well positioned to identify, track, and control the growing fleet of drones. The wide-area, quality, and secure connectivity provided by mobile networks can enhance the efficiency and effectiveness of drone operations beyond visual line-of-sight range. In this article, we elaborate how the drone ecosystem can benefit from mobile technologies, summarize key capabilities required by drone applications, and analyze the service requirements on mobile networks. We present field trial results collected in LTE-Advanced networks to gain insights into the capabilities of the current 4G+ networks for connected drones and share our vision on how 5G networks can further support diversified drone applications.

cs.NI↗

arXiv2018

AAANE: Attention-based Adversarial Autoencoder for Multi-scale Network Embedding

Network embedding represents nodes in a continuous vector space and preserves structure information from the Network. Existing methods usually adopt a "one-size-fits-all" approach when concerning multi-scale structure information, such as first- and second-order proximity of nodes, ignoring the fact that different scales play different roles in the embedding learning. In this paper, we propose an Attention-based Adversarial Autoencoder Network Embedding(AAANE) framework, which promotes the collaboration of different scales and lets them vote for robust representations. The proposed AAANE consists of two components: 1) Attention-based autoencoder effectively capture the highly non-linear network structure, which can de-emphasize irrelevant scales during training. 2) An adversarial regularization guides the autoencoder learn robust representations by matching the posterior distribution of the latent embeddings to given prior distribution. This is the first attempt to introduce attention mechanisms to multi-scale network embedding. Experimental results on real-world networks show that our learned attention parameters are different for every network and the proposed approach outperforms existing state-of-the-art approaches for network embedding.

cs.LG↗

arXiv2018

Deep learning based supervised semantic segmentation of Electron Cryo-Subtomograms

Cellular Electron Cryo-Tomography (CECT) is a powerful imaging technique for the 3D visualization of cellular structure and organization at submolecular resolution. It enables analyzing the native structures of macromolecular complexes and their spatial organization inside single cells. However, due to the high degree of structural complexity and practical imaging limitations, systematic macromolecular structural recovery inside CECT images remains challenging. Particularly, the recovery of a macromolecule is likely to be biased by its neighbor structures due to the high molecular crowding. To reduce the bias, here we introduce a novel 3D convolutional neural network inspired by Fully Convolutional Network and Encoder-Decoder Architecture for the supervised segmentation of macromolecules of interest in subtomograms. The tests of our models on realistically simulated CECT data demonstrate that our new approach has significantly improved segmentation performance compared to our baseline approach. Also, we demonstrate that the proposed model has generalization ability to segment new structures that do not exist in training data.

q-bio.QM↗

arXiv2018

Feature Decomposition Based Saliency Detection in Electron Cryo-Tomograms

Electron Cryo-Tomography (ECT) allows 3D visualization of subcellular structures at the submolecular resolution in close to the native state. However, due to the high degree of structural complexity and imaging limits, the automatic segmentation of cellular components from ECT images is very difficult. To complement and speed up existing segmentation methods, it is desirable to develop a generic cell component segmentation method that is 1) not specific to particular types of cellular components, 2) able to segment unknown cellular components, 3) fully unsupervised and does not rely on the availability of training data. As an important step towards this goal, in this paper, we propose a saliency detection method that computes the likelihood that a subregion in a tomogram stands out from the background. Our method consists of four steps: supervoxel over-segmentation, feature extraction, feature matrix decomposition, and computation of saliency. The method produces a distribution map that represents the regions' saliency in tomograms. Our experiments show that our method can successfully label most salient regions detected by a human observer, and able to filter out regions not containing cellular components. Therefore, our method can remove the majority of the background region, and significantly speed up the subsequent processing of segmentation and recognition of cellular components captured by ECT.

q-bio.QM↗

arXiv2018

Model compression for faster structural separation of macromolecules captured by Cellular Electron Cryo-Tomography

Electron Cryo-Tomography (ECT) enables 3D visualization of macromolecule structure inside single cells. Macromolecule classification approaches based on convolutional neural networks (CNN) were developed to separate millions of macromolecules captured from ECT systematically. However, given the fast accumulation of ECT data, it will soon become necessary to use CNN models to efficiently and accurately separate substantially more macromolecules at the prediction stage, which requires additional computational costs. To speed up the prediction, we compress classification models into compact neural networks with little in accuracy for deployment. Specifically, we propose to perform model compression through knowledge distillation. Firstly, a complex teacher network is trained to generate soft labels with better classification feasibility followed by training of customized student networks with simple architectures using the soft label to compress model complexity. Our tests demonstrate that our compressed models significantly reduce the number of parameters and time cost while maintaining similar classification accuracy.

q-bio.QM↗

arXiv2017

A convolutional autoencoder approach for mining features in cellular electron cryo-tomograms and weakly supervised coarse segmentation

Cellular electron cryo-tomography enables the 3D visualization of cellular organization in the near-native state and at submolecular resolution. However, the contents of cellular tomograms are often complex, making it difficult to automatically isolate different in situ cellular components. In this paper, we propose a convolutional autoencoder-based unsupervised approach to provide a coarse grouping of 3D small subvolumes extracted from tomograms. We demonstrate that the autoencoder can be used for efficient and coarse characterization of features of macromolecular complexes and surfaces, such as membranes. In addition, the autoencoder can be used to detect non-cellular features related to sample preparation and data collection, such as carbon edges from the grid and tomogram boundaries. The autoencoder is also able to detect patterns that may indicate spatial interactions between cellular components. Furthermore, we demonstrate that our autoencoder can be used for weakly supervised semantic segmentation of cellular components, requiring a very small amount of manual annotation.

q-bio.QM↗

arXiv2017

Performance Optimization and Parallelization of a Parabolic Equation Solver in Computational Ocean Acoustics on Modern Many-core Computer

As one of open-source codes widely used in computational ocean acoustics, FOR3D can provide a very good estimate for underwater acoustic propagation. In this paper, we propose a performance optimization and parallelization to speed up the running of FOR3D. We utilized a variety of methods to enhance the entire performance, such as using a multi-threaded programming model to exploit the potential capability of the many-core node of high-performance computing (HPC) system, tuning compile options, using efficient tuned mathematical library and utilizing vectorization optimization instruction. In addition, we extended the application from single-frequency calculation to multi-frequency calculation successfully by using OpenMP+MPI hybrid programming techniques on the mainstream HPC platform. A detailed performance evaluation was performed and the results showed that the proposed parallelization obtained good accelerated effect of 25.77X when testing a typical three-dimensional medium-sized case on Tianhe-2 supercomputer. It also showed that the tuned parallel version has a weak-scalability. The speed of calculation of underwater sound field can be greatly improved by the strategy mentioned in this paper. The method used in this paper is not only applicable to other similar computing models in computational ocean acoustics but also a guideline of performance enhancement for scientific and engineering application running on modern many-core-computing platform.

cs.MS↗

arXiv2017

Faster Convergence of a Randomized Coordinate Descent Method for Linearly Constrained Optimization Problems

The problem of minimizing a separable convex function under linearly coupled constraints arises from various application domains such as economic systems, distributed control, and network flow. The main challenge for solving this problem is that the size of data is very large, which makes usual gradient-based methods infeasible. Recently, Necoara, Nesterov, and Glineur [Journal of Optimization Theory and Applications, 173 (2017) 227-2254] proposed an efficient randomized coordinate descent method to solve this type of optimization problems and presented an appealing convergence analysis. In this paper, we develop new techniques to analyze the convergence of such algorithms, which are able to greatly improve the results presented there. This refined result is achieved by extending Nesterov's second technique developed by Nesterov [SIAM J. Optim. 22 (2012) 341-362] to the general optimization problems with linearly coupled constraints. A novel technique in our analysis is to establish the basis vectors for the subspace of the linearly constraints.

math.OC↗

arXiv2017

Plum pudding random medium model of biological tissue toward remote microscopy from spectroscopic light scattering

Biological tissue has a complex structure and exhibits rich spectroscopic behavior. There is \emph{no} tissue model up to now able to account for the observed spectroscopy of tissue light scattering and its anisotropy. Here we present, \emph{for the first time}, a plum pudding random medium (PPRM) model for biological tissue which succinctly describes tissue as a superposition of distinctive scattering structures (plum) embedded inside a fractal continuous medium of background refractive index fluctuation (pudding). PPRM faithfully reproduces the wavelength dependence of tissue light scattering and attributes the "anomalous" trend in the anisotropy to the plum and the powerlaw dependence of the reduced scattering coefficient to the fractal scattering pudding. Most importantly, PPRM opens up a novel venue of quantifying the tissue architecture and microscopic structures on average from macroscopic probing of the bulk with scattered light alone without tissue excision. We demonstrate this potential by visualizing the fine microscopic structural alterations in breast tissue (adipose, glandular, fibrocystic, fibroadenoma, and ductal carcinoma) deduced from noncontact spectroscopic measurement.

physics.bio-ph↗

arXiv2017

Improved deep learning based macromolecules structure classification from electron cryo tomograms

Cellular processes are governed by macromolecular complexes inside the cell. Study of the native structures of macromolecular complexes has been extremely difficult due to lack of data. With recent breakthroughs in Cellular electron cryo tomography (CECT) 3D imaging technology, it is now possible for researchers to gain accesses to fully study and understand the macromolecular structures single cells. However, systematic recovery of macromolecular structures from CECT is very difficult due to high degree of structural complexity and practical imaging limitations. Specifically, we proposed a deep learning based image classification approach for large-scale systematic macromolecular structure separation from CECT data. However, our previous work was only a very initial step towards exploration of the full potential of deep learning based macromolecule separation. In this paper, we focus on improving classification performance by proposing three newly designed individual CNN models: an extended version of (Deep Small Receptive Field) DSRF3D, donated as DSRF3D-v2, a 3D residual block based neural network, named as RB3D and a convolutional 3D(C3D) based model, CB3D. We compare them with our previously developed model (DSRF3D) on 12 datasets with different SNRs and tilt angle ranges. The experiments show that our new models achieved significantly higher classification accuracies. The accuracies are not only higher than 0.9 on normal datasets, but also demonstrate potentials to operate on datasets with high levels of noises and missing wedge effects presented.

q-bio.QM↗

arXiv2017

Experience enrichment based task independent reward model

For most reinforcement learning approaches, the learning is performed by maximizing an accumulative reward that is expectedly and manually defined for specific tasks. However, in real world, rewards are emergent phenomena from the complex interactions between agents and environments. In this paper, we propose an implicit generic reward model for reinforcement learning. Unlike those rewards that are manually defined for specific tasks, such implicit reward is task independent. It only comes from the deviation from the agents' previous experiences.

cs.AI↗