arXiv subjects

Kai Ma

Publications and source records attributed to Kai Ma.

At least 73 records · Page 4Linked to original sources

arXiv2022

Dense Cross-Query-and-Support Attention Weighted Mask Aggregation for Few-Shot Segmentation

Research into Few-shot Semantic Segmentation (FSS) has attracted great attention, with the goal to segment target objects in a query image given only a few annotated support images of the target class. A key to this challenging task is to fully utilize the information in the support images by exploiting fine-grained correlations between the query and support images. However, most existing approaches either compressed the support information into a few class-wise prototypes, or used partial support information (e.g., only foreground) at the pixel level, causing non-negligible information loss. In this paper, we propose Dense pixel-wise Cross-query-and-support Attention weighted Mask Aggregation (DCAMA), where both foreground and background support information are fully exploited via multi-level pixel-wise correlations between paired query and support features. Implemented with the scaled dot-product attention in the Transformer architecture, DCAMA treats every query pixel as a token, computes its similarities with all support pixels, and predicts its segmentation label as an additive aggregation of all the support pixels' labels -- weighted by the similarities. Based on the unique formulation of DCAMA, we further propose efficient and effective one-pass inference for n-shot segmentation, where pixels of all support images are collected for the mask aggregation at once. Experiments show that our DCAMA significantly advances the state of the art on standard FSS benchmarks of PASCAL-5i, COCO-20i, and FSS-1000, e.g., with 3.1%, 9.7%, and 3.6% absolute improvements in 1-shot mIoU over previous best records. Ablative studies also verify the design DCAMA.

cs.CV↗

arXiv2022

Fast Distributed Stochastic Scheduling for A Multi-Energy Industrial Park

The multi-energy management framework of industrial parks advocates energy conversion and scheduling, which takes full advantage of the compensation and temporal availability of multiple energy. However, how to exploit elastic loads and compensate inelastic loads to match multiple generators and storage is still a key problem under the uncertainty of demand and supply. To solve the issue, the energy management problem is constructed as a stochastic optimization problem. The optimization aims are to minimize the time-averaged energy cost and improve the energy efficiency while respecting the energy constraints. To achieve the distributed implementation in real time without knowing any priori knowledge of underlying stochastic process, a distributed stochastic gradient algorithm based on dual decomposition and a fast scheme are proposed. The numerical results based on real data show that the industrial park, by adopting the proposed algorithm, can achieve social welfare maximization asymptotically.

eess.SY↗

arXiv2022

Mono-$γ$ Production of a Vector Dark Matter at Future $e^+e^-$ Collider

Associated production of a dark particle and a photon, represented as a mono-$γ$ event, is a promising channel to probe particle contents and dynamics in the dark sector. In this paper we study properties of the mono-$γ$ production of a vector dark matter at future $e^+e^-$ colliders. The photon-like and Pauli operators, as well as triple gauge bosons interactions involving the dark matter, are considered in the framework of Effective Field Theory. We show that, comparing to the Pauli operator, the triple gauge bosons couplings are much more interesting at high energy collider. Beam polarization effects are also analyzed, and we show that the experimental sensitivities can not be enhanced significantly because of the smaller luminosity.

hep-ph↗

arXiv2022

Hadron Collider Probes of the Quartic Couplings of Gluons to the Photon and $Z$ Boson

We explore the experimental sensitivities of measuring the $gg \rightarrow Z γ$ process at the LHC to the dimension-8 quartic couplings of gluon pairs to the $Z$ boson and photon, in addition to comparing them with the analogous sensitivities in the $gg \to γγ$ process. These processes can both receive contributions from 4 different CP-conserving dimension-8 operators with distinct Lorentz structures that contain a pair of gluon field strengths, $\hat G^a_{μν}$, and a pair of electroweak SU(2) gauge field strengths, $W^i_{μν}$, as well as 4 similar operators containing a pair of $\hat G^a_{μν}$ and a pair of U(1) gauge field strengths, $B_{μν}$. We calculate the scattering angular distributions for $gg \rightarrow Z γ$ and the $Z \to \bar f f$ decay angular distributions for these 4 Lorentz structures, as well as the Standard Model background. We analyze the sensitivity of ATLAS measurements of the $Z(\to \ell^+\ell^-, \bar νν, \bar q q)γ$ final states with integrated luminosities up to 139 fb$^{-1}$ at $\sqrt{s} = 13$ TeV, showing that they exclude values $\lesssim 2$ TeV for the dimension-8 operator scales, and compare the $Z γ$ sensitivity with that of an ATLAS measurement of the $γγ$ final state. We present combined $Z γ$ and $γγ$ constraints on the scales of dimension-8 SMEFT operators and $γγ$ constraints on the nonlinearity scale of the Born-Infeld extension of the Standard Model. We also estimate the sensitivities to dimension-8 operators of experiments at possible future proton-proton colliders with centre-of-mass energies of 25, 50 and 100 TeV, and discuss possible measurements of the $Z$ spin and angular correlations.

hep-ph↗

arXiv2022

Learning Shape Priors by Pairwise Comparison for Robust Semantic Segmentation

Semantic segmentation is important in medical image analysis. Inspired by the strong ability of traditional image analysis techniques in capturing shape priors and inter-subject similarity, many deep learning (DL) models have been recently proposed to exploit such prior information and achieved robust performance. However, these two types of important prior information are usually studied separately in existing models. In this paper, we propose a novel DL model to model both type of priors within a single framework. Specifically, we introduce an extra encoder into the classic encoder-decoder structure to form a Siamese structure for the encoders, where one of them takes a target image as input (the image-encoder), and the other concatenates a template image and its foreground regions as input (the template-encoder). The template-encoder encodes the shape priors and appearance characteristics of each foreground class in the template image. A cosine similarity based attention module is proposed to fuse the information from both encoders, to utilize both types of prior information encoded by the template-encoder and model the inter-subject similarity for each foreground class. Extensive experiments on two public datasets demonstrate that our proposed method can produce superior performance to competing methods.

cs.CV↗

arXiv2022

DFTR: Depth-supervised Fusion Transformer for Salient Object Detection

Automated salient object detection (SOD) plays an increasingly crucial role in many computer vision applications. By reformulating the depth information as supervision rather than as input, depth-supervised convolutional neural networks (CNN) have achieved promising results on both RGB and RGB-D SOD scenarios with the merits of no requirements for extra depth networks and depth inputs in the inference stage. This paper, for the first time, seeks to expand the applicability of depth supervision to the Transformer architecture. Specifically, we develop a Depth-supervised Fusion TRansformer (DFTR), to further improve the accuracy of both RGB and RGB-D SOD. The proposed DFTR involves three primary features: 1) DFTR, to the best of our knowledge, is the first pure Transformer-based model for depth-supervised SOD; 2) A multi-scale feature aggregation (MFA) module is proposed to fully exploit the multi-scale features encoded by the Swin Transformer in a coarse-to-fine manner; 3) To enable bidirectional information flow across different streams of features, a novel multi-stage feature fusion (MFF) module is further integrated into our DFTR with the emphasis on salient regions at different network learning stages. We extensively evaluate the proposed DFTR on ten benchmarking datasets. Experimental results show that our DFTR consistently outperforms the existing state-of-the-art methods for both RGB and RGB-D SOD tasks. The code and model will be made publicly available.

cs.CV↗

arXiv2022

Polarization and Correlation Effects in Lepton Flavor Violated Decays Induced by Axion-Like Particle

Lepton flavor violated processes are strongly suppressed in the Standard Model, but can be sizable in some extended models. The current experimental bounds on decay modes $\ell'^{\pm} \to a \ell^{\pm}$ are much weaker than the other processes because of the huge irreducible backgrounds $\ell'^{\pm} \to \ell^{\pm} \barν_{\ell} ν_{\ell'}$. In this paper, we study polarization effects of both the signals and backgrounds. We find that signals and backgrounds have distinctive polarization effects, and for the irreducible backgrounds both longitudinal and transverse polarization effects survive when the relative momentum of the two neutrinos are integrated out. At low energy $e^+e^-$ collider, for instance the Belle II experiment, leptons are generated in pair but with no net polarization. However, we show that polarization correlation of the lepton pair is a useful observable for probing the signal. More interestingly, the polarization correlation depends on product of scalar and pseudo-scalar couplings, and hence are sensitive to their relative sign. Because kinematical reconstruction or tagging is not necessary, number of the available signal events increases significantly.

hep-ph↗

arXiv2022

Stochastic Gradient-based Fast Distributed Multi-Energy Management for an Industrial Park with Temporally-Coupled Constraints

Contemporary industrial parks are challenged by the growing concerns about high cost and low efficiency of energy supply. Moreover, in the case of uncertain supply/demand, how to mobilize delay-tolerant elastic loads and compensate real-time inelastic loads to match multi-energy generation/storage and minimize energy cost is a key issue. Since energy management is hardly to be implemented offline without knowing statistical information of random variables, this paper presents a systematic online energy cost minimization framework to fulfill the complementary utilization of multi-energy with time-varying generation, demand and price. Specifically to achieve charging/discharging constraints due to storage and short-term energy balancing, a fast distributed algorithm based on stochastic gradient with two-timescale implementation is proposed to ensure online implementation. To reduce the peak loads, an incentive mechanism is implemented by estimating users' willingness to shift. Analytical results on parameter setting are also given to guarantee feasibility and optimality of the proposed design. Numerical results show that when the bid-ask spread of electricity is small enough, the proposed algorithm can achieve the close-to-optimal cost asymptotically.

eess.SY↗

arXiv2022

Domain Adaptation Meets Zero-Shot Learning: An Annotation-Efficient Approach to Multi-Modality Medical Image Segmentation

Due to the lack of properly annotated medical data, exploring the generalization capability of the deep model is becoming a public concern. Zero-shot learning (ZSL) has emerged in recent years to equip the deep model with the ability to recognize unseen classes. However, existing studies mainly focus on natural images, which utilize linguistic models to extract auxiliary information for ZSL. It is impractical to apply the natural image ZSL solutions directly to medical images, since the medical terminology is very domain-specific, and it is not easy to acquire linguistic models for the medical terminology. In this work, we propose a new paradigm of ZSL specifically for medical images utilizing cross-modality information. We make three main contributions with the proposed paradigm. First, we extract the prior knowledge about the segmentation targets, called relation prototypes, from the prior model and then propose a cross-modality adaptation module to inherit the prototypes to the zero-shot model. Second, we propose a relation prototype awareness module to make the zero-shot model aware of information contained in the prototypes. Last but not least, we develop an inheritance attention module to recalibrate the relation prototypes to enhance the inheritance process. The proposed framework is evaluated on two public cross-modality datasets including a cardiac dataset and an abdominal dataset. Extensive experiments show that the proposed framework significantly outperforms the state of the arts.

cs.CV↗

arXiv2022

All-Around Real Label Supervision: Cyclic Prototype Consistency Learning for Semi-supervised Medical Image Segmentation

Semi-supervised learning has substantially advanced medical image segmentation since it alleviates the heavy burden of acquiring the costly expert-examined annotations. Especially, the consistency-based approaches have attracted more attention for their superior performance, wherein the real labels are only utilized to supervise their paired images via supervised loss while the unlabeled images are exploited by enforcing the perturbation-based \textit{"unsupervised"} consistency without explicit guidance from those real labels. However, intuitively, the expert-examined real labels contain more reliable supervision signals. Observing this, we ask an unexplored but interesting question: can we exploit the unlabeled data via explicit real label supervision for semi-supervised training? To this end, we discard the previous perturbation-based consistency but absorb the essence of non-parametric prototype learning. Based on the prototypical network, we then propose a novel cyclic prototype consistency learning (CPCL) framework, which is constructed by a labeled-to-unlabeled (L2U) prototypical forward process and an unlabeled-to-labeled (U2L) backward process. Such two processes synergistically enhance the segmentation network by encouraging more discriminative and compact features. In this way, our framework turns previous \textit{"unsupervised"} consistency into new \textit{"supervised"} consistency, obtaining the \textit{"all-around real label supervision"} property of our method. Extensive experiments on brain tumor segmentation from MRI and kidney segmentation from CT images show that our CPCL can effectively exploit the unlabeled data and outperform other state-of-the-art semi-supervised medical image segmentation methods.

eess.IV↗

arXiv2022

Conquering Data Variations in Resolution: A Slice-Aware Multi-Branch Decoder Network

Fully convolutional neural networks have made promising progress in joint liver and liver tumor segmentation. Instead of following the debates over 2D versus 3D networks (for example, pursuing the balance between large-scale 2D pretraining and 3D context), in this paper, we novelly identify the wide variation in the ratio between intra- and inter-slice resolutions as a crucial obstacle to the performance. To tackle the mismatch between the intra- and inter-slice information, we propose a slice-aware 2.5D network that emphasizes extracting discriminative features utilizing not only in-plane semantics but also out-of-plane coherence for each separate slice. Specifically, we present a slice-wise multi-input multi-output architecture to instantiate such a design paradigm, which contains a Multi-Branch Decoder (MD) with a Slice-centric Attention Block (SAB) for learning slice-specific features and a Densely Connected Dice (DCD) loss to regularize the inter-slice predictions to be coherent and continuous. Based on the aforementioned innovations, we achieve state-of-the-art results on the MICCAI 2017 Liver Tumor Segmentation (LiTS) dataset. Besides, we also test our model on the ISBI 2019 Segmentation of THoracic Organs at Risk (SegTHOR) dataset, and the result proves the robustness and generalizability of the proposed method in other segmentation tasks.

eess.IV↗

arXiv2022

Simultaneous Alignment and Surface Regression Using Hybrid 2D-3D Networks for 3D Coherent Layer Segmentation of Retina OCT Images

Automated surface segmentation of retinal layer is important and challenging in analyzing optical coherence tomography (OCT). Recently, many deep learning based methods have been developed for this task and yield remarkable performance. However, due to large spatial gap and potential mismatch between the B-scans of OCT data, all of them are based on 2D segmentation of individual B-scans, which may loss the continuity information across the B-scans. In addition, 3D surface of the retina layers can provide more diagnostic information, which is crucial in quantitative image analysis. In this study, a novel framework based on hybrid 2D-3D convolutional neural networks (CNNs) is proposed to obtain continuous 3D retinal layer surfaces from OCT. The 2D features of individual B-scans are extracted by an encoder consisting of 2D convolutions. These 2D features are then used to produce the alignment displacement field and layer segmentation by two 3D decoders, which are coupled via a spatial transformer module. The entire framework is trained end-to-end. To the best of our knowledge, this is the first study that attempts 3D retinal layer segmentation in volumetric OCT images based on CNNs. Experiments on a publicly available dataset show that our framework achieves superior results to state-of-the-art 2D methods in terms of both layer segmentation accuracy and cross-B-scan 3D continuity, thus offering more clinical values than previous works.

eess.IV↗

arXiv2022

Bidirectional Pricing and Demand Response for Nanogrids with HVAC Systems

Owing to the fluctuant renewable generation and power demand, the energy surplus or deficit in each nanogrid is embodied differently across time. To stimulate local renewable energy consumption and minimize the long-term energy cost, some issues still remain to be explored: when and how the energy demand and bidirectional trading prices are scheduled considering personal comfort preferences and environmental factors. For this purpose, the demand response and two-way pricing problems concurrently for nanogrids and a public monitoring entity (PME) are studied with exploiting the large potential thermal elastic ability of heating, ventilation and air-conditioning (HVAC) units. Different from nanogrids, in terms of minimizing time-average costs, PME aims to set reasonable prices and optimize profits by trading with nanogrids and the main grid bi-directionally. In particular, such bilevel energy management problem is formulated as a stochastic form in a long-term horizon. Since there are uncertain system parameters, time-coupled queue constraints and the interplay of bilevel decision-making, it is challenging to solve the formulated problems. To this end, we derive a form of relaxation based on Lyapunov optimization technique to make the energy management problem tractable without forecasting the related system parameters. The transaction between nanogrids and PME is captured by a one-leader and multi-follower Stackelberg game framework. Then, theoretical analysis of the existence and uniqueness of Stackelberg equilibrium (SE) is developed based on the proposed game property. Following that, we devise an optimization algorithm to reach the SE with less information exchange. Numerical experiments validate the effectiveness of the proposed approach.

cs.GT↗

arXiv2022

Asynchronous Decentralized Federated Learning for Collaborative Fault Diagnosis of PV Stations

Due to the different losses caused by various photovoltaic (PV) array faults, accurate diagnosis of fault types is becoming increasingly important. Compared with a single one, multiple PV stations collect sufficient fault samples, but their data is not allowed to be shared directly due to potential conflicts of interest. Therefore, federated learning can be exploited to train a collaborative fault diagnosis model. However, the modeling efficiency is seriously affected by the model update mechanism since each PV station has a different computing capability and amount of data. Moreover, for the safe and stable operation of the PV system, the robustness of collaborative modeling must be guaranteed rather than simply being processed on a central server. To address these challenges, a novel asynchronous decentralized federated learning (ADFL) framework is proposed. Each PV station not only trains its local model but also participates in collaborative fault diagnosis by exchanging model parameters to improve the generalization without losing accuracy. The global model is aggregated distributedly to avoid central node failure. By designing the asynchronous update scheme, the communication overhead and training time are greatly reduced. Both the experiments and numerical simulations are carried out to verify the effectiveness of the proposed method.

cs.LG↗

arXiv2022

Energy Management Based on Multi-Agent Deep Reinforcement Learning for A Multi-Energy Industrial Park

Owing to large industrial energy consumption, industrial production has brought a huge burden to the grid in terms of renewable energy access and power supply. Due to the coupling of multiple energy sources and the uncertainty of renewable energy and demand, centralized methods require large calculation and coordination overhead. Thus, this paper proposes a multi-energy management framework achieved by decentralized execution and centralized training for an industrial park. The energy management problem is formulated as a partially-observable Markov decision process, which is intractable by dynamic programming due to the lack of the prior knowledge of the underlying stochastic process. The objective is to minimize long-term energy costs while ensuring the demand of users. To solve this issue and improve the calculation speed, a novel multi-agent deep reinforcement learning algorithm is proposed, which contains the following key points: counterfactual baseline for facilitating contributing agents to learn better policies, soft actor-critic for improving robustness and exploring optimal solutions. A novel reward is designed by Lagrange multiplier method to ensure the capacity constraints of energy storage. In addition, considering that the increase in the number of agents leads to performance degradation due to large observation spaces, an attention mechanism is introduced to enhance the stability of policy and enable agents to focus on important energy-related information, which improves the exploration efficiency of soft actor-critic. Numerical results based on actual data verify the performance of the proposed algorithm with high scalability, indicating that the industrial park can minimize energy costs under different demands.

eess.SY↗

arXiv2021

Alleviating Noisy-label Effects in Image Classification via Probability Transition Matrix

Deep-learning-based image classification frameworks often suffer from the noisy label problem caused by the inter-observer variation. Recent studies employed learning-to-learn paradigms (e.g., Co-teaching and JoCoR) to filter the samples with noisy labels from the training set. However, most of them use a simple cross-entropy loss as the criterion for noisy label identification. The hard samples, which are beneficial for classifier learning, are often mistakenly treated as noises in such a setting since both the hard samples and ones with noisy labels lead to a relatively larger loss value than the easy cases. In this paper, we propose a plugin module, namely noise ignoring block (NIB), consisting of a probability transition matrix and an inter-class correlation (IC) loss, to separate the hard samples from the mislabeled ones, and further boost the accuracy of image classification network trained with noisy labels. Concretely, our IC loss is calculated as Kullback-Leibler divergence between the network prediction and the accumulative soft label generated by the probability transition matrix. Such that, with the lower value of IC loss, the hard cases can be easily distinguished from mislabeled cases. Extensive experiments are conducted on natural and medical image datasets (CIFAR-10 and ISIC 2019). The experimental results show that our NIB module consistently improves the performances of the state-of-the-art robust training methods.

cs.CV↗

arXiv2021

A Unified Framework for Generalized Low-Shot Medical Image Segmentation with Scarce Data

Medical image segmentation has achieved remarkable advancements using deep neural networks (DNNs). However, DNNs often need big amounts of data and annotations for training, both of which can be difficult and costly to obtain. In this work, we propose a unified framework for generalized low-shot (one- and few-shot) medical image segmentation based on distance metric learning (DML). Unlike most existing methods which only deal with the lack of annotations while assuming abundance of data, our framework works with extreme scarcity of both, which is ideal for rare diseases. Via DML, the framework learns a multimodal mixture representation for each category, and performs dense predictions based on cosine distances between the pixels' deep embeddings and the category representations. The multimodal representations effectively utilize the inter-subject similarities and intraclass variations to overcome overfitting due to extremely limited data. In addition, we propose adaptive mixing coefficients for the multimodal mixture distributions to adaptively emphasize the modes better suited to the current input. The representations are implicitly embedded as weights of the fc layer, such that the cosine distances can be computed efficiently via forward propagation. In our experiments on brain MRI and abdominal CT datasets, the proposed framework achieves superior performances for low-shot segmentation towards standard DNN-based (3D U-Net) and classical registration-based (ANTs) methods, e.g., achieving mean Dice coefficients of 81%/69% for brain tissue/abdominal multiorgan segmentation using a single training sample, as compared to 52%/31% and 72%/35% by the U-Net and ANTs, respectively.

cs.CV↗

arXiv2021

Unsupervised Representation Learning Meets Pseudo-Label Supervised Self-Distillation: A New Approach to Rare Disease Classification

Rare diseases are characterized by low prevalence and are often chronically debilitating or life-threatening. Imaging-based classification of rare diseases is challenging due to the severe shortage in training examples. Few-shot learning (FSL) methods tackle this challenge by extracting generalizable prior knowledge from a large base dataset of common diseases and normal controls, and transferring the knowledge to rare diseases. Yet, most existing methods require the base dataset to be labeled and do not make full use of the precious examples of the rare diseases. To this end, we propose in this work a novel hybrid approach to rare disease classification, featuring two key novelties targeted at the above drawbacks. First, we adopt the unsupervised representation learning (URL) based on self-supervising contrastive loss, whereby to eliminate the overhead in labeling the base dataset. Second, we integrate the URL with pseudo-label supervised classification for effective self-distillation of the knowledge about the rare diseases, composing a hybrid approach taking advantages of both unsupervised and (pseudo-) supervised learning on the base dataset. Experimental results on classification of rare skin lesions show that our hybrid approach substantially outperforms existing FSL methods (including those using fully supervised base dataset) for rare disease classification via effective integration of the URL and pseudo-label driven self-distillation, thus establishing a new state of the art.

cs.CV↗