A graph convolutional network with dynamic weight fusion of multi-scale local features for diabetic retinopathy grading.

Diabetic retinopathy (DR) is a serious ocular complication that can pose a serious risk to a patient's vision and overall health. Currently, the automatic grading of DR is mainly using deep learning techniques. However, the lesion information in DR images is complex, variable in shape and size, and randomly distributed in the images, which leads to some shortcomings of the current research methods, i.e., it is difficult to effectively extract the information of these various features, and it is difficult to establish the connection between the lesion information in different regions. To address these shortcomings, we design a multi-scale dynamic fusion (MSDF) module and combine it with graph convolution operations to propose a multi-scale dynamic graph convolutional network (MDGNet) in this paper. MDGNet firstly uses convolution kernels with different sizes to extract features with different shapes and sizes in the lesion regions, and then automatically learns the corresponding weights for feature fusion according to the contribution of different features to model grading. Finally, the graph convolution operation is used to link the lesion features in different regions. As a result, our proposed method can effectively combine local and global features, which is beneficial for the correct DR grading. We evaluate the effectiveness of method on two publicly available datasets, namely APTOS and DDR. Extensive experiments demonstrate that our proposed MDGNet achieves the best grading results on APTOS and DDR, and is more accurate and diverse for the extraction of lesion information.

Probabilistic Contrastive Learning for Long-Tailed Visual Recognition.

Long-tailed distributions frequently emerge in real-world data, where a large number of minority categories contain a limited number of samples. Such imbalance issue considerably impairs the performance of standard supervised learning algorithms, which are mainly designed for balanced training sets. Recent investigations have revealed that supervised contrastive learning exhibits promising potential in alleviating the data imbalance. However, the performance of supervised contrastive learning is plagued by an inherent challenge: it necessitates sufficiently large batches of training data to construct contrastive pairs that cover all categories, yet this requirement is difficult to meet in the context of class-imbalanced data. To overcome this obstacle, we propose a novel probabilistic contrastive (ProCo) learning algorithm that estimates the data distribution of the samples from each class in the feature space, and samples contrastive pairs accordingly. In fact, estimating the distributions of all classes using features in a small batch, particularly for imbalanced data, is not feasible. Our key idea is to introduce a reasonable and simple assumption that the normalized features in contrastive learning follow a mixture of von Mises-Fisher (vMF) distributions on unit space, which brings two-fold benefits. First, the distribution parameters can be estimated using only the first sample moment, which can be efficiently computed in an online manner across different batches. Second, based on the estimated distribution, the vMF distribution allows us to sample an infinite number of contrastive pairs and derive a closed form of the expected contrastive loss for efficient optimization. Other than long-tailed problems, ProCo can be directly applied to semi-supervised learning by generating pseudo-labels for unlabeled data, which can subsequently be utilized to estimate the distribution of the samples inversely. Theoretically, we analyze the error bound of ProCo. Empirically, extensive experimental results on supervised/semi-supervised visual recognition and object detection tasks demonstrate that ProCo consistently outperforms existing methods across various datasets.

Domain Adaptation via Prompt Learning.

Unsupervised domain adaptation (UDA) aims to adapt models learned from a well-annotated source domain to a target domain, where only unlabeled samples are given. Current UDA approaches learn domain-invariant features by aligning source and target feature spaces through statistical discrepancy minimization or adversarial training. However, these constraints could lead to the distortion of semantic feature structures and loss of class discriminability. In this article, we introduce a novel prompt learning paradigm for UDA, named domain adaptation via prompt learning (DAPrompt). In contrast to prior works, our approach learns the underlying label distribution for target domain rather than aligning domains. The main idea is to embed domain information into prompts, a form of representation generated from natural language, which is then used to perform classification. This domain information is shared only by images from the same domain, thereby dynamically adapting the classifier according to each domain. By adopting this paradigm, we show that our model not only outperforms previous methods on several cross-domain benchmarks but also is very efficient to train and easy to implement.

Hundreds Guide Millions: Adaptive Offline Reinforcement Learning With Expert Guidance.

Offline reinforcement learning (RL) optimizes the policy on a previously collected dataset without any interactions with the environment, yet usually suffers from the distributional shift problem. To mitigate this issue, a typical solution is to impose a policy constraint on a policy improvement objective. However, existing methods generally adopt a "one-size-fits-all" practice, i.e., keeping only a single improvement-constraint balance for all the samples in a mini-batch or even the entire offline dataset. In this work, we argue that different samples should be treated with different policy constraint intensities. Based on this idea, a novel plug-in approach named guided offline RL (GORL) is proposed. GORL employs a guiding network, along with only a few expert demonstrations, to adaptively determine the relative importance of the policy improvement and policy constraint for every sample. We theoretically prove that the guidance provided by our method is rational and near-optimal. Extensive experiments on various environments suggest that GORL can be easily installed on most offline RL algorithms with statistically significant performance improvements.

MLP-based classification of COVID-19 and skin diseases.

Recent years have witnessed a growing interest in neural network-based medical image classification methods, which have demonstrated remarkable performance in this field. Typically, convolutional neural network (CNN) architectures have been commonly employed to extract local features. However, the transformer, a newly emerged architecture, has gained popularity due to its ability to explore the relevance of remote elements in an image through a self-attention mechanism. Despite this, it is crucial to establish not only local connectivity but also remote relationships between lesion features and capture the overall image structure to improve image classification accuracy. Therefore, to tackle the aforementioned issues, this paper proposes a network based on multilayer perceptrons (MLPs) that can learn the local features of medical images on the one hand and capture the overall feature information in both spatial and channel dimensions on the other hand, thus utilizing image features effectively. This paper has been extensively validated on COVID19-CT dataset and ISIC 2018 dataset, and the results show that the method in this paper is more competitive and has higher performance in medical image classification compared with existing methods. This shows that the use of MLP to capture image features and establish connections between lesions is expected to provide novel ideas for medical image classification tasks in the future.

Glance and Focus Networks for Dynamic Visual Recognition.

Spatial redundancy widely exists in visual recognition tasks, i.e., discriminative features in an image or video frame usually correspond to only a subset of pixels, while the remaining regions are irrelevant to the task at hand. Therefore, static models which process all the pixels with an equal amount of computation result in considerable redundancy in terms of time and space consumption. In this paper, we formulate the image recognition problem as a sequential coarse-to-fine feature learning process, mimicking the human visual system. Specifically, the proposed Glance and Focus Network (GFNet) first extracts a quick global representation of the input image at a low resolution scale, and then strategically attends to a series of salient (small) regions to learn finer features. The sequential process naturally facilitates adaptive inference at test time, as it can be terminated once the model is sufficiently confident about its prediction, avoiding further redundant computation. It is worth noting that the problem of locating discriminant regions in our model is formulated as a reinforcement learning task, thus requiring no additional manual annotations other than classification labels. GFNet is general and flexible as it is compatible with any off-the-shelf backbone models (such as MobileNets, EfficientNets and TSM), which can be conveniently deployed as the feature extractor. Extensive experiments on a variety of image classification and video recognition tasks and with various backbone models demonstrate the remarkable efficiency of our method. For example, it reduces the average latency of the highly efficient MobileNet-V3 on an iPhone XS Max by 1.3x without sacrificing accuracy. Code and pre-trained models are available at https://github.com/blackfeather-wang/GFNet-Pytorch.

Meta-Reinforcement Learning With Dynamic Adaptiveness Distillation.

Deep reinforcement learning is confronted with problems of sampling inefficiency and poor task migration capability. Meta-reinforcement learning (meta-RL) enables meta-learners to utilize the task-solving skills trained on similar tasks and quickly adapt to new tasks. However, meta-RL methods lack enough queries toward the relationship between task-agnostic exploitation of data and task-related knowledge introduced by latent context, limiting their effectiveness and generalization ability. In this article, we develop an algorithm for off-policy meta-RL that can provide the meta-learners with self-oriented cognition toward how they adapt to the family of tasks. In our approach, we perform dynamic task-adaptiveness distillation to describe how the meta-learners adjust the exploration strategy in the meta-training process. Our approach also enables the meta-learners to balance the influence of task-agnostic self-oriented adaption and task-related information through latent context reorganization. In our experiments, our method achieves 10%-20% higher asymptotic reward than probabilistic embeddings for actor-critic RL (PEARL).

Input-to-State Stability of Nonlinear Impulsive Systems Subjects to Actuator Saturation and External Disturbance.

This article mainly explores the local input-to-state stability (LISS) property of a class of nonlinear systems via a saturated control strategy, where both the external disturbance and impulsive disturbance being fully considered. In terms of the Lyapunov method and inequality techniques, some sufficient conditions under which the system can be made LISS are proposed, and the elastic constraint relationship among saturated control gain, rate coefficients, external disturbance, and domain of initial value is revealed. Moreover, the optimization design procedures are provided with the hope of obtaining the estimates of admissible external disturbance and domain of initial value as large as possible, where the corresponding saturated control law can be designed by solving LMI -based conditions. In the absence of an external disturbance, the locally exponential stability (LES) property can also be presented with a set of more relaxed conditions. Finally, two examples are presented to reveal the validity of the obtained results.

The Hot Strip Mill Scheduling Problem With Uncertainty: Robust Optimization Models and Solution Approaches.

In this article, we focus on a biobjective hot strip mill (HSM) scheduling problem arising in the steel industry. Besides the conventional objective regarding penalty costs, we have also considered minimizing the total starting times of rolling operations in order to reduce the energy consumption for slab reheating. The problem is complicated by the inevitable uncertainty in rolling processing times, which means deterministic scheduling models will be ineffective. To obtain robust production schedules with satisfactory performance under all possible conditions, we apply the robust optimization (RO) approach to model and solve the scheduling problem. First, an RO model and an equivalent mixed-integer linear programming model are constructed to describe the HSM scheduling problem with uncertainty. Then, we devise an improved Benders' decomposition algorithm to solve the RO model and obtain exactly optimal solutions. Next, for coping with large-sized instances, a multiobjective particle swarm optimization algorithm with an embedded local search strategy is proposed to handle the biobjective scheduling problem and find the set of Pareto-optimal solutions. Finally, we conduct extensive computational tests to verify the proposed algorithms. Results show that the exact algorithm is effective for relatively small instances and the metaheuristic algorithm can achieve satisfactory solution quality for both small- and large-sized instances of the problem.

Exploration With Task Information for Meta Reinforcement Learning.

Meta reinforcement learning (meta-RL) is a promising technique for fast task adaptation by leveraging prior knowledge from previous tasks. Recently, context-based meta-RL has been proposed to improve data efficiency by applying a principled framework, dividing the learning procedure into task inference and task execution. However, the task information is not adequately leveraged in this approach, thus leading to inefficient exploration. To address this problem, we propose a novel context-based meta-RL framework with an improved exploration mechanism. For the existing exploration and execution problem in context-based meta-RL, we propose a novel objective that employs two exploration terms to encourage better exploration in action and task embedding space, respectively. The first term pushes for improving the diversity of task inference, while the second term, named action information, works as sharing or hiding task information in different exploration stages. We divide the meta-training procedure into task-independent exploration and task-relevant exploration stages according to the utilization of action information. By decoupling task inference and task execution and proposing the respective optimization objectives in the two exploration stages, we can efficiently learn policy and task inference networks. We compare our algorithm with several popular meta-RL methods on MuJoco benchmarks with both dense and sparse reward settings. The empirical results show that our method significantly outperforms baselines on the benchmarks in terms of sample efficiency and task performance.

Finite-time stability of state-dependent delayed systems and application to coupled neural networks.

Finite-time stability and stabilization problems of state-dependent delayed systems are studied in this paper. Different from discrete delays and time-dependent delays which can be well estimated over time, the information of state-dependent delays is usually hard to be estimated, especially when states are unknown or unmeasurable. To guarantee the stability of state-dependent delayed systems in the framework of finite time, a Razumikhin-type inequality is used, following which estimations on the settling time and the region of attraction are proposed. Moreover, the relationship between the variation speed of state-dependent delays and the size of the region of attraction is proposed. Then as an application of the theoretical result, finite-time stabilization is studied for a set of nonlinear coupled neural networks involving state-dependent transmission delay, where the design of memoryless finite-time controllers is addressed. Two numerical examples are given to show the effectiveness of the proposed results.

Multi-scale feature progressive fusion network for remote sensing image change detection.

Presently, research on deep learning-based change detection (CD) methods has become a hot topic. In particular, feature pyramid networks (FPNs) are widely used in CD tasks to gradually fuse semantic features. However, existing FPN-based CD methods do not correctly detect the complete change region and cannot accurately locate the boundaries of the change region. To solve these problems, a new Multi-Scale Feature Progressive Fusion Network (MFPF-Net) is proposed, which consists of three innovative modules: Layer Feature Fusion Module (LFFM), Multi-Scale Feature Aggregation Module (MSFA), and Multi-Scale Feature Distribution Module (MSFD). Specifically, we first concatenate the features of each layer extracted from the bi-temporal images with their difference maps, and the resulting change maps fuse richer semantic information while effectively representing change regions. Then, the obtained change maps of each layer are directly aggregated, which improves the effective communication and full fusion of feature maps in CD while avoiding the interference of indirect information. Finally, the aggregated feature maps are layered again by pooling and convolution operations, and then a feature fusion strategy with a pyramid structure is used, with layers fused from low to high, to obtain richer contextual information, so that each layer of the layered feature maps has original semantic information and semantic features of other layers. We conducted comprehensive experiments on three publicly available benchmark datasets, CDD, LEVIR-CD, and WHU-CD to verify the effectiveness of the method, and the experimental results show that the method in this paper outperforms other comparative methods.

Optimization of a molten iron scheduling problem with uncertain processing time using variable neighborhood search algorithm.

Punctuality of the steel-making scheduling is important to save steel production costs, but the processing time of the pretreatment process, which connects the iron- and steel-making stages, is usually uncertain. This paper presents a distributionally robust iron-steel allocation (DRISA) model to obtain a robust scheduling plan, where the distribution of the pretreatment time vector is assumed to belong to an ambiguity set which contains all the distributions with given first and second moments. This model aims to minimize the production objective by determining the iron-steel allocation and the completion time of each charge, while the constraints should hold with a certain probability under the worst-case distribution. To solve problems in large-scale efficiently, a variable neighborhood algorithm is developed to obtain a near-optimal solution in a short time. Experiments based on actual production data demonstrate its efficiency. Results also show the robustness of the DRISA model, i.e., the adjustment and delay of the robust schedule derived from the DRISA model are less than the nominal one.

Temporal-Spatial Causal Interpretations for Vision-Based Reinforcement Learning.

Deep reinforcement learning (RL) agents are becoming increasingly proficient in a range of complex control tasks. However, the agent's behavior is usually difficult to interpret due to the introduction of black-box function, making it difficult to acquire the trust of users. Although there have been some interesting interpretation methods for vision-based RL, most of them cannot uncover temporal causal information, raising questions about their reliability. To address this problem, we present a temporal-spatial causal interpretation (TSCI) model to understand the agent's long-term behavior, which is essential for sequential decision-making. TSCI model builds on the formulation of temporal causality, which reflects the temporal causal relations between sequential observations and decisions of RL agent. Then a separate causal discovery network is employed to identify temporal-spatial causal features, which are constrained to satisfy the temporal causality. TSCI model is applicable to recurrent agents and can be used to discover causal features with high efficiency once trained. The empirical results show that TSCI model can produce high-resolution and sharp attention masks to highlight task-relevant temporal-spatial information that constitutes most evidence about how vision-based RL agents make sequential decisions. In addition, we further demonstrate that our method is able to provide valuable causal interpretations for vision-based RL agents from the temporal perspective.

A Hybrid Artificial Immune-Simulated Annealing Algorithm for Multiroute Job Shop Scheduling Problem With Continuous Limited Output Buffers.

In this article, we study the multiroute job shop scheduling problem with continuous-limited output buffers (MRJSP-CLOBs). In contrast to the standard job shop scheduling problem (JSP), continuous-limited output buffers render the commonly used graph-based approaches inapplicable, and the multiroute issue further increases computational complexity. To this end, we formulate MRJSP-CLOB as a mixed-integer linear program (MILP), which is typically NP-hard. Then, we extend the critical block in the JSP by utilizing the no-time-gap relationship and design a new neighborhood structure. Furthermore, we propose a hybrid artificial immune-simulated annealing algorithm (AIA-SA) by sharing iterations and integrating a random infeasible solution repairing algorithm with a new SA acceptance rule, which enables individuals to share information and increases the robustness of the corresponding SA parameters. Finally, the AIA-SA is compared with CPLEX and state-of-the-art algorithms on MRJSP-CLOB with different sizes. Experiments for large-sized instances demonstrate that our algorithm requires less than 3% computing time of the CPLEX, while being faster and more accurate than the other algorithms.

Regularizing Deep Networks With Semantic Data Augmentation.

Data augmentation is widely known as a simple yet surprisingly effective technique for regularizing deep networks. Conventional data augmentation schemes, e.g., flipping, translation or rotation, are low-level, data-independent and class-agnostic operations, leading to limited diversity for augmented samples. To this end, we propose a novel semantic data augmentation algorithm to complement traditional approaches. The proposed method is inspired by the intriguing property that deep networks are effective in learning linearized features, i.e., certain directions in the deep feature space correspond to meaningful semantic transformations, e.g., changing the background or view angle of an object. Based on this observation, translating training samples along many such directions in the feature space can effectively augment the dataset for more diversity. To implement this idea, we first introduce a sampling based method to obtain semantically meaningful directions efficiently. Then, an upper bound of the expected cross-entropy (CE) loss on the augmented training set is derived by assuming the number of augmented samples goes to infinity, yielding a highly efficient algorithm. In fact, we show that the proposed implicit semantic data augmentation (ISDA) algorithm amounts to minimizing a novel robust CE loss, which adds minimal extra computational cost to a normal training procedure. In addition to supervised learning, ISDA can be applied to semi-supervised learning tasks under the consistency regularization framework, where ISDA amounts to minimizing the upper bound of the expected KL-divergence between the augmented features and the original features. Although being simple, ISDA consistently improves the generalization performance of popular deep models (e.g., ResNets and DenseNets) on a variety of datasets, i.e., CIFAR-10, CIFAR-100, SVHN, ImageNet, and Cityscapes. Code for reproducing our results is available at https://github.com/blackfeather-wang/ISDA-for-Deep-Networks.

Attention-Based Meta-Reinforcement Learning for Tracking Control of AUV With Time-Varying Dynamics.

Reinforcement learning (RL) is a promising technique for designing a model-free controller by interacting with the environment. Several researchers have applied RL to autonomous underwater vehicles (AUVs) for motion control, such as trajectory tracking. However, the existing RL-based controller usually assumes that the unknown AUV dynamics keep invariant during the operation period, limiting its further application in the complex underwater environment. In this article, a novel meta-RL-based control scheme is proposed for trajectory tracking control of AUV in the presence of unknown and time-varying dynamics. To this end, we divide the tracking task for AUV with time-varying dynamics into multiple specific tasks with fixed time-varying dynamics, to which we apply meta-RL for training to distill the general control policy. The obtained control policy can transfer to the testing phase with high adaptability. Inspired by the line-of-sight (LOS) tracking rule, we formulate each specific task as a Markov decision process (MDP) with a well-designed state and reward function. Furthermore, a novel policy network with an attention module is proposed to extract the hidden information of AUV dynamics. The simulation environment with time-varying dynamics is established, and the simulation results reveal the effectiveness of our proposed method.

Self-Supervised Discovering of Interpretable Features for Reinforcement Learning.

Deep reinforcement learning (RL) has recently led to many breakthroughs on a range of complex control tasks. However, the agent's decision-making process is generally not transparent. The lack of interpretability hinders the applicability of RL in safety-critical scenarios. While several methods have attempted to interpret vision-based RL, most come without detailed explanation for the agent's behavior. In this paper, we propose a self-supervised interpretable framework, which can discover interpretable features to enable easy understanding of RL agents even for non-experts. Specifically, a self-supervised interpretable network (SSINet) is employed to produce fine-grained attention masks for highlighting task-relevant information, which constitutes most evidence for the agent's decisions. We verify and evaluate our method on several Atari 2600 games as well as Duckietown, which is a challenging self-driving car simulator environment. The results show that our method renders empirical evidences about how the agent makes decisions and why the agent performs well or badly, especially when transferred to novel scenes. Overall, our method provides valuable insight into the internal decision-making process of vision-based RL. In addition, our method does not use any external labelled data, and thus demonstrates the possibility to learn high-quality mask through a self-supervised manner, which may shed light on new paradigms for label-free vision learning such as self-supervised segmentation and detection.

Finite-Time Synchronization for Delayed Complex Dynamical Networks With Synchronizing or Desynchronizing Impulses.

In this article, the finite-time synchronization problem of delayed complex dynamical networks (CDNs) with impulses is studied, where two types of impulses, namely, synchronizing impulses and desynchronizing impulses, are fully considered, respectively. Since the existence of impulses makes the discontinuity of the states, which means that the classical result for finite-time stability is inapplicable in such a case, the key challenge is how to guarantee the finite-time stability and estimate the settling time in impulse sense. We apply impulsive control theory and finite-time stability theory to CDNs and establish some sufficient conditions for finite-time synchronization, where two kinds of memory controllers are designed for synchronizing impulses and desynchronizing impulses, respectively. Moreover, the upper bounds for settling time of synchronization, which depends on the impulse sequences, are effectively estimated. It shows that the synchronizing impulses can shorten the settling time of synchronization; conversely, the desynchronizing impulses can delay it. Finally, the theoretical analysis is verified by two simulation examples.

Dynamic Neural Networks: A Survey.

Dynamic neural network is an emerging research topic in deep learning. Compared to static models which have fixed computational graphs and parameters at the inference stage, dynamic networks can adapt their structures or parameters to different inputs, leading to notable advantages in terms of accuracy, computational efficiency, adaptiveness, etc. In this survey, we comprehensively review this rapidly developing area by dividing dynamic networks into three main categories: 1) sample-wise dynamic models that process each sample with data-dependent architectures or parameters; 2) spatial-wise dynamic networks that conduct adaptive computation with respect to different spatial locations of image data; and 3) temporal-wise dynamic models that perform adaptive inference along the temporal dimension for sequential data such as videos and texts. The important research problems of dynamic networks, e.g., architecture design, decision making scheme, optimization technique and applications, are reviewed systematically. Finally, we discuss the open problems in this field together with interesting future research directions.