arXiv subjects

Wenhao Li

Publications and source records attributed to Wenhao Li.

251 records · Page 14Linked to original sources

arXiv2021

Interactive Medical Image Segmentation with Self-Adaptive Confidence Calibration

Medical image segmentation is one of the fundamental problems for artificial intelligence-based clinical decision systems. Current automatic medical image segmentation methods are often failed to meet clinical requirements. As such, a series of interactive segmentation algorithms are proposed to utilize expert correction information. However, existing methods suffer from some segmentation refining failure problems after long-term interactions and some cost problems from expert annotation, which hinder clinical applications. This paper proposes an interactive segmentation framework, called interactive MEdical segmentation with self-adaptive Confidence CAlibration (MECCA), by introducing the corrective action evaluation, which combines the action-based confidence learning and multi-agent reinforcement learning (MARL). The evaluation is established through a novel action-based confidence network, and the corrective actions are obtained from MARL. Based on the confidential information, a self-adaptive reward function is designed to provide more detailed feedback, and a simulated label generation mechanism is proposed on unsupervised data to reduce over-reliance on labeled data. Experimental results on various medical image datasets have shown the significant performance of the proposed algorithm.

cs.CV↗

arXiv2021

HMRL: Hyper-Meta Learning for Sparse Reward Reinforcement Learning Problem

In spite of the success of existing meta reinforcement learning methods, they still have difficulty in learning a meta policy effectively for RL problems with sparse reward. In this respect, we develop a novel meta reinforcement learning framework called Hyper-Meta RL(HMRL), for sparse reward RL problems. It is consisted with three modules including the cross-environment meta state embedding module which constructs a common meta state space to adapt to different environments; the meta state based environment-specific meta reward shaping which effectively extends the original sparse reward trajectory by cross-environmental knowledge complementarity and as a consequence the meta policy achieves better generalization and efficiency with the shaped meta reward. Experiments with sparse-reward environments show the superiority of HMRL on both transferability and policy learning efficiency.

cs.LG↗

arXiv2021

CCPM: A Chinese Classical Poetry Matching Dataset

Poetry is one of the most important art forms of human languages. Recently many studies have focused on incorporating some linguistic features of poetry, such as style and sentiment, into its understanding or generation system. However, there is no focus on understanding or evaluating the semantics of poetry. Therefore, we propose a novel task to assess a model's semantic understanding of poetry by poem matching. Specifically, this task requires the model to select one line of Chinese classical poetry among four candidates according to the modern Chinese translation of a line of poetry. To construct this dataset, we first obtain a set of parallel data of Chinese classical poetry and modern Chinese translation. Then we retrieve similar lines of poetry with the lines in a poetry corpus as negative choices. We name the dataset Chinese Classical Poetry Matching Dataset (CCPM) and release it at https://github.com/THUNLP-AIPoet/CCPM. We hope this dataset can further enhance the study on incorporating deep semantics into the understanding and generation system of Chinese classical poetry. We also preliminarily run two variants of BERT on this dataset as the baselines for this dataset.

cs.CL↗

arXiv2021

Structured Diversification Emergence via Reinforced Organization Control and Hierarchical Consensus Learning

When solving a complex task, humans will spontaneously form teams and to complete different parts of the whole task, respectively. Meanwhile, the cooperation between teammates will improve efficiency. However, for current cooperative MARL methods, the cooperation team is constructed through either heuristics or end-to-end blackbox optimization. In order to improve the efficiency of cooperation and exploration, we propose a structured diversification emergence MARL framework named {\sc{Rochico}} based on reinforced organization control and hierarchical consensus learning. {\sc{Rochico}} first learns an adaptive grouping policy through the organization control module, which is established by independent multi-agent reinforcement learning. Further, the hierarchical consensus module based on the hierarchical intentions with consensus constraint is introduced after team formation. Simultaneously, utilizing the hierarchical consensus module and a self-supervised intrinsic reward enhanced decision module, the proposed cooperative MARL algorithm {\sc{Rochico}} can output the final diversified multi-agent cooperative policy. All three modules are organically combined to promote the structured diversification emergence. Comparative experiments on four large-scale cooperation tasks show that {\sc{Rochico}} is significantly better than the current SOTA algorithms in terms of exploration efficiency and cooperation strength.

cs.LG↗

arXiv2021

DHLink: A Microservice Platform supporting Rapid Application Development and Secure Real-time Data Sharing in Digital Health

Digital health applications that leverage multiple sources of patient data for insights to patients' behaviours or disease symptoms as well as remote patient monitoring, nudging and treatments are becoming increasingly popular in various medical practices and research. One common issue among these applications is that they are generally based on project-specific solutions and developed from scratch. Such application development fashion results in large amounts of repetitive effort, for example, in building study specific websites and mobile frontends, deploying customised infrastructures, and collecting data that may have already been collected in other studies and projects. What is worse, the data collected, and functions built cannot be easily reused by other applications. In this paper, we present an event-driven microservice platform, namely DHLink, to address this issue. DHLink securely links existing digital health applications of different projects, facilitates real-time data sharing, and supports rapid application development by reusing data and functions of existing digital health applications. In addition, comes with DHLink, a set of highly generic and reusable microservices is provided, which allows developers to rapidly create a typical above-mentioned digital health application by only developing the core algorithms. Two use cases outlined in this paper have shown the use of DHLink and the set of microservices for application collaboration and new application development to be efficient and practical.

cs.DC↗

arXiv2020

Using Information to Amplify Competition

I characterize the consumer-optimal market segmentation in competitive markets where multiple firms selling differentiated products to consumers with unit demand. This segmentation is public---in that each firm observes the same market segments---and takes a simple form: in each market segment, there is a dominant firm favored by all consumers in that segment. By segmenting the market, all but the dominant firm maximally compete to poach the consumer's business, setting price to equal marginal cost. Information, thus, is being used to amplify competition. This segmentation simultaneously generates an efficient allocation and delivers to each firm its minimax profit.

econ.GN↗

arXiv2020

Malicious Requests Detection with Improved Bidirectional Long Short-term Memory Neural Networks

Detecting and intercepting malicious requests are one of the most widely used ways against attacks in the network security. Most existing detecting approaches, including matching blacklist characters and machine learning algorithms have all shown to be vulnerable to sophisticated attacks. To address the above issues, a more general and rigorous detection method is required. In this paper, we formulate the problem of detecting malicious requests as a temporal sequence classification problem, and propose a novel deep learning model namely Convolutional Neural Network-Bidirectional Long Short-term Memory-Convolutional Neural Network (CNN-BiLSTM-CNN). By connecting the shadow and deep feature maps of the convolutional layers, the malicious feature extracting ability is improved on more detailed functionality. Experimental results on HTTP dataset CSIC 2010 have demonstrated the effectiveness of the proposed method when compared with the state-of-the-arts.

cs.LG↗

arXiv2020

MixPoet: Diverse Poetry Generation via Learning Controllable Mixed Latent Space

As an essential step towards computer creativity, automatic poetry generation has gained increasing attention these years. Though recent neural models make prominent progress in some criteria of poetry quality, generated poems still suffer from the problem of poor diversity. Related literature researches show that different factors, such as life experience, historical background, etc., would influence composition styles of poets, which considerably contributes to the high diversity of human-authored poetry. Inspired by this, we propose MixPoet, a novel model that absorbs multiple factors to create various styles and promote diversity. Based on a semi-supervised variational autoencoder, our model disentangles the latent space into some subspaces, with each conditioned on one influence factor by adversarial training. In this way, the model learns a controllable latent variable to capture and mix generalized factor-related properties. Different factor mixtures lead to diverse styles and hence further differentiate generated poems from each other. Experiment results on Chinese poetry demonstrate that MixPoet improves both diversity and quality against three state-of-the-art models.

cs.CL↗

arXiv2020

Learning Structured Communication for Multi-agent Reinforcement Learning

This work explores the large-scale multi-agent communication mechanism under a multi-agent reinforcement learning (MARL) setting. We summarize the general categories of topology for communication structures in MARL literature, which are often manually specified. Then we propose a novel framework termed as Learning Structured Communication (LSC) by using a more flexible and efficient communication topology. Our framework allows for adaptive agent grouping to form different hierarchical formations over episodes, which is generated by an auxiliary task combined with a hierarchical routing protocol. Given each formed topology, a hierarchical graph neural network is learned to enable effective message information generation and propagation among inter- and intra-group communications. In contrast to existing communication mechanisms, our method has an explicit while learnable design for hierarchical communication. Experiments on challenging tasks show the proposed LSC enjoys high communication efficiency, scalability, and global cooperation capability.

cs.LG↗

arXiv2020

A continuous contact force model for impact analysis in multibody dynamics

A new continuous contact force model for contacting problems with regular or irregular contacting surfaces and energy dissipations in multibody systems is presented and discussed in this work. The model is developed according to Hertz law and a hysteresis damping force is introduced for modeling the energy dissipation during the contact process. As it is almost impossible to obtain an analytical solution based on the system dynamic equation, an approximate dynamic equation for the collision system is proposed, achieving a good approximation of the system dynamic equation. An approximate function between deformation velocity and deformation is founded on the approximate dynamic equation, then it is utilized to calculate the energy loss due to the damping force. The model is established through modifying the original formula of the hysteresis damping parameter derived by combining the energy balance and the law of conservation of linear momentum. Numerical results of five different continuous contact models reveal the capability of our new model as well as the effect of the geometry of the contacting surfaces on the dynamic system response.

physics.class-ph↗

arXiv2019

Iteratively-Refined Interactive 3D Medical Image Segmentation with Multi-Agent Reinforcement Learning

Existing automatic 3D image segmentation methods usually fail to meet the clinic use. Many studies have explored an interactive strategy to improve the image segmentation performance by iteratively incorporating user hints. However, the dynamic process for successive interactions is largely ignored. We here propose to model the dynamic process of iterative interactive image segmentation as a Markov decision process (MDP) and solve it with reinforcement learning (RL). Unfortunately, it is intractable to use single-agent RL for voxel-wise prediction due to the large exploration space. To reduce the exploration space to a tractable size, we treat each voxel as an agent with a shared voxel-level behavior strategy so that it can be solved with multi-agent reinforcement learning. An additional advantage of this multi-agent model is to capture the dependency among voxels for segmentation task. Meanwhile, to enrich the information of previous segmentations, we reserve the prediction uncertainty in the state space of MDP and derive an adjustment action space leading to a more precise and finer segmentation. In addition, to improve the efficiency of exploration, we design a relative cross-entropy gain-based reward to update the policy in a constrained direction. Experimental results on various medical datasets have shown that our method significantly outperforms existing state-of-the-art methods, with the advantage of fewer interactions and a faster convergence.

cs.CV↗

arXiv2019

Bright magnetic dipole radiation from two-dimensional lead-halide perovskites

Light-matter interactions in semiconductor systems are uniformly treated within the electric dipole (ED) approximation, as multipolar interactions are considered "forbidden". Here, we demonstrate that this approximation inadequately describes light emission in novel two-dimensional hybrid organic-inorganic perovskite materials (2D HOIPs) --- a class of solution processable layered semiconductor with promising optoelectronic properties. Consequently, photoluminescence (PL) spectra become strongly dependent on the experimental geometry, a fact that is often overlooked, though critical for correct optical characterization of materials. Using energy-momentum and time-resolved spectroscopies, we experimentally demonstrate that low-energy sideband emission in 2D HOIPs exhibits a highly unusual, multipolar polarization and angle dependence. Using combined electromagnetic and quantum-mechanical analyses, we attribute this radiation pattern to an out-of-plane oriented magnetic dipole transition arising from the 2D character of the excited and ground state orbitals. Symmetry arguments point toward the presence of significant inversion symmetry-breaking mechanisms that are currently under great debate. These results provide a new perspective on the origins of unexpected sideband emission in HOIPs, clarify discrepancies in previous literature, and generally challenge the paradigm of ED-dominated light-matter interactions in novel optoelectronic materials.

cond-mat.mtrl-sci↗

arXiv2017

Research on two-dimensional traffic flow model based on psychological field theory

In this paper, the influence of fan-shaped buffer zone on the performance of the toll plaza is researched. A two-dimensional traffic flow model and a comprehensive evaluation model based on mechanical model and psychological field are established. The traffic flow model is simulated by creating coordinate system. We first establish queue theory model to analyze vehicles when entering toll plaza. Then, a two-dimensional steadily car-following model is established based on psychological field for the analysis of vehicles when leaving toll plaza. According to psychological field theory, we analyze the force condition of each vehicle. The force of each vehicle is contributed by the vehicles in its observation area and obstacles. By projecting these vehicles and obstacles via the equipotential line in the psychological field, the influence on the value and direction acceleration of following vehicles is obtained. Consequently, the changes of each vehicle's speed and position are obtained as well. Next, we establish simulation based on the states of vehicles and make the rules of vehicle state-changing. By simulating the system, we obtain the throughput of the toll plaza's input and output. Then we obtained the bearing pressure on the road by the max throughput and the demand of the roads. Using the number of cars in per unit area as the safety factor. Then a comprehensive evaluation model is established based on bearing pressure on the road, cost and safety factor.

physics.soc-ph↗

arXiv2017

Research on ruin probability of risk model based on AR(1) series

In this text, we establish the risk model based on AR(1) series and propose the basic model which has a dependent structure under intensity of claim number. Considering some properties of the risk model, we take advantage of newton iteration method to figure out the adjustment coefficient and estimate the exponential upper bound of ruin probability. This is significant to refine the research of ruin theory. As a result, our theory will help develop insurance industry stably.

q-fin.RM↗

arXiv2017

Hybrid optical pumping of K and Rb atoms in a paraffin coated vapor cell

Dynamic hybrid optical pumping effects with a radio-frequency-field-driven nonlinear magneto-optical rotation (RF NMOR) scheme are studied in a dual-species paraffin coated vapor cell. By pumping K atoms and probing $^{87}$Rb atoms, we achieve an intrinsic magnetic resonance linewidth of 3 Hz and the observed resonance is immune to power broadening and light-shift effects. Such operation scheme shows favorable prospects for atomic magnetometry applications.

physics.atom-ph↗

arXiv2017

An Architecture for Establishing Legal Semantic Workflows in the Context of Integrated Law Enforcement

Traditionally the integration of data from multiple sources is done on an ad-hoc basis for each analysis scenario and application. This is a solution that is inflexible, incurs in high costs, leads to "silos" that prevent sharing data across different agencies or tasks, and is unable to cope with the modern environment, where workflows, tasks, and priorities frequently change. Operating within the Data to Decision Cooperative Research Centre (D2D CRC), the authors are currently involved in the Integrated Law Enforcement Project, which has the goal of developing a federated data platform that will enable the execution of integrated analytics on data accessed from different external and internal sources, thereby providing effective support to an investigator or analyst working to evaluate evidence and manage lines of inquiries in the investigation. Technical solutions should also operate ethically, in compliance with the law, and subject to good governance principles.

cs.CY↗

arXiv2016

Characterization of high-temperature performance of cesium vapor cells with anti-relaxation coating

Vapor cells with antirelaxation coating are widely used in modern atomic physics experiments due to the coating's ability to maintain the atoms' spin polarization during wall collisions. We characterize the performance of vapor cells with different coating materials by measuring longitudinal spin relaxation and vapor density at temperatures up to 95°C. We found that the spin-projection-noise-limited sensitivity for atomic magnetometers with such cells improves with temperature, which demonstrates the potential of antirelaxation coated cells in applications of future high-sensitivity magnetometers.

physics.atom-ph↗