arXiv subjects

Chao Chen

Publications and source records attributed to Chao Chen.

At least 847 records · Page 47Linked to original sources

arXiv2020

Semantic Segmentation for Partially Occluded Apple Trees Based on Deep Learning

Fruit tree pruning and fruit thinning require a powerful vision system that can provide high resolution segmentation of the fruit trees and their branches. However, recent works only consider the dormant season, where there are minimal occlusions on the branches or fit a polynomial curve to reconstruct branch shape and hence, losing information about branch thickness. In this work, we apply two state-of-the-art supervised learning models U-Net and DeepLabv3, and a conditional Generative Adversarial Network Pix2Pix (with and without the discriminator) to segment partially occluded 2D-open-V apple trees. Binary accuracy, Mean IoU, Boundary F1 score and Occluded branch recall were used to evaluate the performances of the models. DeepLabv3 outperforms the other models at Binary accuracy, Mean IoU and Boundary F1 score, but is surpassed by Pix2Pix (without discriminator) and U-Net in Occluded branch recall. We define two difficulty indices to quantify the difficulty of the task: (1) Occlusion Difficulty Index and (2) Depth Difficulty Index. We analyze the worst 10 images in both difficulty indices by means of Branch Recall and Occluded Branch Recall. U-Net outperforms the other two models in the current metrics. On the other hand, Pix2Pix (without discriminator) provides more information on branch paths, which are not reflected by the metrics. This highlights the need for more specific metrics on recovering occluded information. Furthermore, this shows the usefulness of image-transfer networks for hallucination behind occlusions. Future work is required to further enhance the models to recover more information from occlusions such that this technology can be applied to automating agricultural tasks in a commercial environment.

cs.CV↗

arXiv2020

Dirac-Born-Infeld realization of sound speed resonance mechanism for primordial black holes

We present a concrete realization of the sound speed resonance (SSR) mechanism for primordial black hole (PBH) formation within a specific model of Dirac-Born-Infeld (DBI) inflation. We perform a perturbative approach to phenomenologically construct such a viable DBI inflation model that involves the nonoscillating stage and the oscillating stage, with a type of specific forms of the warp factor and the potential. We show that the continuous but nonsmooth conjunction of sound speed between two stages does not yield manifest effects on the phenomenology of SSR, and thus, our model gives rise to the same PBH mass spectrum as the original predictions of SSR. Additionally, we also demonstrate that the violation of adiabaticity of the Mukhanov-Sasaki equation does not affect the comoving curvature perturbation after Hubble crossing in the nonresonant region. Making use of observational data, we derive various cosmological constraints on the parameter space. Our analyses show that the predicted tensor-to-scalar ratio is typically small, while the amplitude of primordial non-Gaussianity can meet with cosmic microwave background bounds, and additionally, the consistency relation for single-field slow-roll inflation is softly violated in our case due to the small sound speed variations.

astro-ph.CO↗

arXiv2020

Attention-Guided Discriminative Region Localization and Label Distribution Learning for Bone Age Assessment

Bone age assessment (BAA) is clinically important as it can be used to diagnose endocrine and metabolic disorders during child development. Existing deep learning based methods for classifying bone age use the global image as input, or exploit local information by annotating extra bounding boxes or key points. However, training with the global image underutilizes discriminative local information, while providing extra annotations is expensive and subjective. In this paper, we propose an attention-guided approach to automatically localize the discriminative regions for BAA without any extra annotations. Specifically, we first train a classification model to learn the attention maps of the discriminative regions, finding the hand region, the most discriminative region (the carpal bones), and the next most discriminative region (the metacarpal bones). Guided by those attention maps, we then crop the informative local regions from the original image and aggregate different regions for BAA. Instead of taking BAA as a general regression task, which is suboptimal due to the label ambiguity problem in the age label space, we propose using joint age distribution learning and expectation regression, which makes use of the ordinal relationship among hand images with different individual ages and leads to more robust age estimation. Extensive experiments are conducted on the RSNA pediatric bone age data set. Using no training annotations, our method achieves competitive results compared with existing state-of-the-art semi-automatic deep learning-based methods that require manual annotation. Code is available at https: //github.com/chenchao666/Bone-Age-Assessment.

cs.CV↗

arXiv2020

Learn distributed GAN with Temporary Discriminators

In this work, we propose a method for training distributed GAN with sequential temporary discriminators. Our proposed method tackles the challenge of training GAN in the federated learning manner: How to update the generator with a flow of temporary discriminators? We apply our proposed method to learn a self-adaptive generator with a series of local discriminators from multiple data centers. We show our design of loss function indeed learns the correct distribution with provable guarantees. The empirical experiments show that our approach is capable of generating synthetic data which is practical for real-world applications such as training a segmentation model.

cs.CV↗

arXiv2020

DRWR: A Differentiable Renderer without Rendering for Unsupervised 3D Structure Learning from Silhouette Images

Differentiable renderers have been used successfully for unsupervised 3D structure learning from 2D images because they can bridge the gap between 3D and 2D. To optimize 3D shape parameters, current renderers rely on pixel-wise losses between rendered images of 3D reconstructions and ground truth images from corresponding viewpoints. Hence they require interpolation of the recovered 3D structure at each pixel, visibility handling, and optionally evaluating a shading model. In contrast, here we propose a Differentiable Renderer Without Rendering (DRWR) that omits these steps. DRWR only relies on a simple but effective loss that evaluates how well the projections of reconstructed 3D point clouds cover the ground truth object silhouette. Specifically, DRWR employs a smooth silhouette loss to pull the projection of each individual 3D point inside the object silhouette, and a structure-aware repulsion loss to push each pair of projections that fall inside the silhouette far away from each other. Although we omit surface interpolation, visibility handling, and shading, our results demonstrate that DRWR achieves state-of-the-art accuracies under widely used benchmarks, outperforming previous methods both qualitatively and quantitatively. In addition, our training times are significantly lower due to the simplicity of DRWR.

cs.CV↗

arXiv2020

End-to-End AI-Based Point-of-Care Diagnosis System for Classifying Respiratory Illnesses and Early Detection of COVID-19

Respiratory symptoms can be a caused by different underlying conditions, and are often caused by viral infections, such as Influenza-like illnesses or other emerging viruses like the Coronavirus. These respiratory viruses, often, have common symptoms, including coughing, high temperature, congested nose, and difficulty breathing. However, early diagnosis of the type of the virus, can be crucial, especially in cases such as the recent COVID-19 pandemic. One of the factors that contributed to the spread of the pandemic, was the late diagnosis or confusing it with regular flu-like symptoms. Science has proved that one of the possible differentiators of the underlying causes of these different respiratory diseases is coughing, which comes in different types and forms. Therefore, a reliable lab-free tool for early and more accurate diagnosis that can differentiate between different respiratory diseases is very much needed. This paper proposes an end-to-end portable system that can record data from patients with symptom, including coughs (voluntary or involuntary) and translate them into health data for diagnosis, and with the aid of machine learning, classify them into different respiratory illnesses, including COVID-19. With the ongoing efforts to stop the spread of the COVID-19 disease everywhere today, and against similar diseases in the future, our proposed low cost and user-friendly solution can play an important part in the early diagnosis.

eess.SP↗

arXiv2020

Local Causal Structure Learning and its Discovery Between Type 2 Diabetes and Bone Mineral Density

Type 2 diabetes (T2DM), one of the most prevalent chronic diseases, affects the glucose metabolism of the human body, which decreases the quantity of life and brings a heavy burden on social medical care. Patients with T2DM are more likely to suffer bone fragility fracture as diabetes affects bone mineral density (BMD). However, the discovery of the determinant factors of BMD in a medical way is expensive and time-consuming. In this paper, we propose a novel algorithm, Prior-Knowledge-driven local Causal structure Learning (PKCL), to discover the underlying causal mechanism between BMD and its factors from the clinical data. Since there exist limited data but redundant prior knowledge for medicine, PKCL adequately utilize the prior knowledge to mine the local causal structure for the target relationship. Combining the medical prior knowledge with the discovered causal relationships, PKCL can achieve more reliable results without long-standing medical statistical experiments. Extensive experiments are conducted on a newly provided clinical data set. The experimental study of PKCL on the data is proved to highly corresponding with existing medical knowledge, which demonstrates the superiority and effectiveness of PKCL. To illustrate the importance of prior knowledge, the result of the algorithm without prior knowledge is also investigated.

q-bio.QM↗

arXiv2020

Synthetic Learning: Learn From Distributed Asynchronized Discriminator GAN Without Sharing Medical Image Data

In this paper, we propose a data privacy-preserving and communication efficient distributed GAN learning framework named Distributed Asynchronized Discriminator GAN (AsynDGAN). Our proposed framework aims to train a central generator learns from distributed discriminator, and use the generated synthetic image solely to train the segmentation model.We validate the proposed framework on the application of health entities learning problem which is known to be privacy sensitive. Our experiments show that our approach: 1) could learn the real image's distribution from multiple datasets without sharing the patient's raw data. 2) is more efficient and requires lower bandwidth than other distributed deep learning methods. 3) achieves higher performance compared to the model trained by one real dataset, and almost the same performance compared to the model trained by all real datasets. 4) has provable guarantees that the generator could learn the distributed distribution in an all important fashion thus is unbiased.

eess.IV↗

arXiv2020

Quasi-one-dimensional metallic conduction channels in exotic ferroelectric topological defects

Ferroelectric topological objects (e.g. vortices, skyrmions) provide a fertile ground for exploring emerging physical properties that could potentially be utilized in future configurable nanoelectronic devices. Here, we demonstrate quasi-one-dimensional metallic high conduction channels along two types of exotic topological defects, i.e. the topological cores of (i) a quadrant vortex domain structure and (ii) a center domain (monopole-like) structure confined in high quality BiFeO3 nanoisland array, abbreviated as the vortex core and the center core. We unveil via phase-field simulations that the superfine (< 3 nm) metallic conduction channels along center cores arise from the screening charge carriers confined at the core whereas the high conductance of vortex cores results from a field-induced twisted state. These conducting channels can be repeatedly and reversibly created and deleted by manipulating the two topological states via an electric field, leading to an apparent electroresistance effect with an on/off ratio higher than 103. These results open up the possibility of utilizing these functional one-dimensional topological objects in high-density nanoelectronic devices such as ultrahigh density nonvolatile memory.

cond-mat.mtrl-sci↗

arXiv2020

Topological Hall effect in single thick SrRuO3 layers induced by defect engineering

The topological Hall effect (THE) has been discovered in ultrathin SrRuO3 (SRO) films, where the interface between the SRO layer and another oxide layer breaks the inversion symmetry resulting in the appearance of THE. Thus, THE only occurs in ultra-thin SRO films of several unit cells. In addition to employing a heterostructure, the inversion symmetry can be broken intrinsically in bulk SRO by introducing defects. In this study THE is observed in 60 nm thick SRO films, in which defects and lattice distortions are introduced by helium ion irradiation. The irradiated SRO films exhibit a pronounced THE in a wide temperature range from 5 K to 80 K. These observations can be attributed to the emergence of Dzyaloshinskii-Moriya interaction as a result of artificial inversion symmetry breaking associated to the lattice defect engineering. The creation and control of the THE in oxide single layers can be realized by ex situ film processing. Therefore, this work provides new insights into the THE and illustrates a promising strategy to design novel spintronics devices.

cond-mat.mtrl-sci↗

arXiv2020

Rigorous Explanation of Inference on Probabilistic Graphical Models

Probabilistic graphical models, such as Markov random fields (MRF), exploit dependencies among random variables to model a rich family of joint probability distributions. Sophisticated inference algorithms, such as belief propagation (BP), can effectively compute the marginal posteriors. Nonetheless, it is still difficult to interpret the inference outcomes for important human decision making. There is no existing method to rigorously attribute the inference outcomes to the contributing factors of the graphical models. Shapley values provide an axiomatic framework, but naively computing or even approximating the values on general graphical models is challenging and less studied. We propose GraphShapley to integrate the decomposability of Shapley values, the structure of MRFs, and the iterative nature of BP inference in a principled way for fast Shapley value computation, that 1) systematically enumerates the important contributions to the Shapley values of the explaining variables without duplicate; 2) incrementally compute the contributions without starting from scratches. We theoretically characterize GraphShapley regarding independence, equal contribution, and additivity. On nine graphs, we demonstrate that GraphShapley provides sensible and practical explanations.

cs.SI↗

arXiv2020

Real-Time Fruit Recognition and Grasping Estimation for Autonomous Apple Harvesting

In this research, a fully neural network based visual perception framework for autonomous apple harvesting is proposed. The proposed framework includes a multi-function neural network for fruit recognition and a Pointnet grasp estimation to determine the proper grasp pose to guide the robotic execution. Fruit recognition takes raw input of RGB images from the RGB-D camera to perform fruit detection and instance segmentation, and Pointnet grasp estimation take point cloud of each fruit as input and output the prediction of grasp pose for each of fruits. The proposed framework is validated by using RGB-D images collected from laboratory and orchard environments, a robotic grasping test in a controlled environment is also included in the experiments. Experimental shows that the proposed framework can accurately localise and estimate the grasp pose for robotic grasping.

cs.CV↗

arXiv2020

Hasse Polynomials of L-functions of Certain Exponential Sums

In this paper, we focus on computing the higher slope Hasse polynomials of L-functions of certain exponential sums associated to the following family of Laurent polynomials $f(x_1,\ldots ,x_{n+1})=\sum_{i=1}^na_i x_{n+1}\left(x_i+\frac{1}{x_i}\right)+a_{n+1} x_{n+1}+\frac{1}{x_{n+1}}$, where $a_i \in \F^*_{q}$, $i=1,2, \ldots, n+1$. We find a simple formula for the Hasse polynomial of the slope one side and study the irreducibility of these Hasse polynomials. We will also provide a simple form of all the higher slope Hasse polynomials for $n=3$, answering an open question of Zhang and Feng.

math.NT↗

arXiv2020

BuildSenSys: Reusing Building Sensing Data for Traffic Prediction with Cross-domain Learning

With the rapid development of smart cities, smart buildings are generating a massive amount of building sensing data by the equipped sensors. Indeed, building sensing data provides a promising way to enrich a series of data-demanding and cost-expensive urban mobile applications. In this paper, we study how to reuse building sensing data to predict traffic volume on nearby roads. Nevertheless, it is non-trivial to achieve accurate prediction on such cross-domain data with two major challenges. First, relationships between building sensing data and traffic data are not unknown as prior, and the spatio-temporal complexities impose more difficulties to uncover the underlying reasons behind the above relationships. Second, it is even more daunting to accurately predict traffic volume with dynamic building-traffic correlations, which are cross-domain, non-linear, and time-varying. To address the above challenges, we design and implement BuildSenSys, a first-of-its-kind system for nearby traffic volume prediction by reusing building sensing data. First, we conduct a comprehensive building-traffic analysis based on multi-source datasets, disclosing how and why building sensing data is correlated with nearby traffic volume. Second, we propose a novel recurrent neural network for traffic volume prediction based on cross-domain learning with two attention mechanisms. Specifically, a cross-domain attention mechanism captures the building-traffic correlations and adaptively extracts the most relevant building sensing data at each predicting step. Then, a temporal attention mechanism is employed to model the temporal dependencies of data across historical time intervals. The extensive experimental studies demonstrate that BuildSenSys outperforms all baseline methods with up to 65.3% accuracy improvement (e.g., 2.2% MAPE) in predicting nearby traffic volume.

eess.SP↗

arXiv2020

Selective Transfer with Reinforced Transfer Network for Partial Domain Adaptation

One crucial aspect of partial domain adaptation (PDA) is how to select the relevant source samples in the shared classes for knowledge transfer. Previous PDA methods tackle this problem by re-weighting the source samples based on their high-level information (deep features). However, since the domain shift between source and target domains, only using the deep features for sample selection is defective. We argue that it is more reasonable to additionally exploit the pixel-level information for PDA problem, as the appearance difference between outlier source classes and target classes is significantly large. In this paper, we propose a reinforced transfer network (RTNet), which utilizes both high-level and pixel-level information for PDA problem. Our RTNet is composed of a reinforced data selector (RDS) based on reinforcement learning (RL), which filters out the outlier source samples, and a domain adaptation model which minimizes the domain discrepancy in the shared label space. Specifically, in the RDS, we design a novel reward based on the reconstruct errors of selected source samples on the target generator, which introduces the pixel-level information to guide the learning of RDS. Besides, we develope a state containing high-level information, which used by the RDS for sample selection. The proposed RDS is a general module, which can be easily integrated into existing DA models to make them fit the PDA situation. Extensive experiments indicate that RTNet can achieve state-of-the-art performance for PDA tasks on several benchmark datasets.

cs.LG↗

arXiv2020

An Algebraic Sparsified Nested Dissection Algorithm Using Low-Rank Approximations

We propose a new algorithm for the fast solution of large, sparse, symmetric positive-definite linear systems, spaND -- sparsified Nested Dissection. It is based on nested dissection, sparsification and low-rank compression. After eliminating all interiors at a given level of the elimination tree, the algorithm sparsifies all separators corresponding to the interiors. This operation reduces the size of the separators by eliminating some degrees of freedom but without introducing any fill-in. This is done at the expense of a small and controllable approximation error. The result is an approximate factorization that can be used as an efficient preconditioner. We then perform several numerical experiments to evaluate this algorithm. We demonstrate that a version using orthogonal factorization and block-diagonal scaling takes fewer CG iterations to converge than previous similar algorithms on various kinds of problems. Furthermore, this algorithm is provably guaranteed to never break down and the matrix stays symmetric positive-definite throughout the process. We evaluate the algorithm on some large problems and show it exhibits near-linear scaling. The factorization time is roughly O(N) and the number of iterations grows slowly with N.

math.NA↗

arXiv2019

Visual Perception and Modelling in Unstructured Orchard for Apple Harvesting Robots

Vision perception and modelling are the essential tasks of robotic harvesting in the unstructured orchard. This paper develops a framework of visual perception and modelling for robotic harvesting of fruits in the orchard environments. The developed framework includes visual perception, scenarios mapping, and fruit modelling. The Visual perception module utilises a deep-learning model to perform multi-purpose visual perception task within the working scenarios; The scenarios mapping module applies OctoMap to represent the multiple classes of objects or elements within the environment; The fruit modelling module estimates the geometry property of objects and estimates the proper access pose of each fruit. The developed framework is implemented and evaluated in the apple orchards. The experiment results show that visual perception and modelling algorithm can accurately detect and localise the fruits, and modelling working scenarios in real orchard environments. The $F_{1}$ score and mean intersection of union of visual perception module on fruit detection and segmentation are 0.833 and 0.852, respectively. The accuracy of the fruit modelling in terms of centre localisation and pose estimation are 0.955 and 0.923, respectively. Overall, an accurate visual perception and modelling algorithm are presented in this paper.

cs.CV↗

arXiv2019

HoMM: Higher-order Moment Matching for Unsupervised Domain Adaptation

Minimizing the discrepancy of feature distributions between different domains is one of the most promising directions in unsupervised domain adaptation. From the perspective of distribution matching, most existing discrepancy-based methods are designed to match the second-order or lower statistics, which however, have limited expression of statistical characteristic for non-Gaussian distributions. In this work, we explore the benefits of using higher-order statistics (mainly refer to third-order and fourth-order statistics) for domain matching. We propose a Higher-order Moment Matching (HoMM) method, and further extend the HoMM into reproducing kernel Hilbert spaces (RKHS). In particular, our proposed HoMM can perform arbitrary-order moment tensor matching, we show that the first-order HoMM is equivalent to Maximum Mean Discrepancy (MMD) and the second-order HoMM is equivalent to Correlation Alignment (CORAL). Moreover, the third-order and the fourth-order moment tensor matching are expected to perform comprehensive domain alignment as higher-order statistics can approximate more complex, non-Gaussian distributions. Besides, we also exploit the pseudo-labeled target samples to learn discriminative representations in the target domain, which further improves the transfer performance. Extensive experiments are conducted, showing that our proposed HoMM consistently outperforms the existing moment matching methods by a large margin. Codes are available at \url{https://github.com/chenchao666/HoMM-Master}

cs.CV↗