Accurate segmentation of gliomas in multi-modal MRI is a critical prerequisite for clinical diagnosis and treatment. Deep learning has made significant progress in automated glioma segmentation. However, existing methods still face challenges in multi-scale feature extraction and multi-modal feature fusion. This paper proposes a Global-Local Collaboration and Task-driven Fusion Network (GLC- TFNet) for glioma segmentation in multi-modal MRIs. To enhance multi-scale feature extraction, a global-local collaboration encoder is proposed, which utilizes a dual-path mechanism to synergistically integrate global contextual information with local details. For multi-modal feature fusion, a task-driven fusion module is designed, which incorporates clinical diagnostic priors as constraints to optimize the contributions of multi-modal features in critical regions. Comprehensive experiments conducted on two publicly released datasets, BraTS2020 and BraTS2021, demon-strate that GLC- TFNet outperforms current state-of-the-art methods. Specifically, it achieves Dice scores of 94.82%,90.08%, and 88.09% for the whole tumor, tumor core, and enhancing tumor on the BraTS2020 dataset, and 92.88%, 90.61 %, and 87.06% on the BraTS2021 dataset, respectively.

Distributed Machine Learning (DML) is proposed to accelerate machine learning model training by utilizing multiple training nodes to train models in parallel. Recent studies apply emerging In-Network Aggregation (INA) to further improve training efficiency by offloading the gradient aggregation process from hosts to programmable switches. However, existing INA solutions neither provide high training performance due to inefficient gradient aggregation with a single switch nor are easily deployed because of modifying the protocol stack in hosts. In this paper, we propose an easily deployable INA-based solution called Hierarchical In-Network Aggregation (HINA) to accelerate DML training process by hierarchically performing multiple aggregations in the Data Center Network (DCN). We formulate the gradient aggregation problem as the Joint Gradient Routing and Sending Rate (JGRSR) problem, which is a Mixed Integer Linear Programming (MILP) problem with high computation complexity. In addition, we propose HINA using progressive rounding and randomized rounding to determine the paths of gradient flows and the sending rates of training nodes to simplify and solve the JGRSR problem. Simulation results show that HINA reduces communication time by 61%-92% and decreases network load by 39%-66%, compared with state-of-the-art solutions.

Distributed Machine Learning (DML) utilizes parallel computation on multiple training nodes to accelerate machine learning model training. Parameter Server (PS) is a typical DML enabler and is widely used in industry and academia. Existing works propose to apply the emerging In-Network Aggregation (INA) technique to improve model training efficiency by offloading the whole gradient aggregation process in PS from hosts to programmable switches. However, existing INA systems may suffer from undetermined model training efficiency and service quality, given that many gradient aggregation processes are still performed by the server under irrational gradient aggregation strategies. In this paper, we propose a Deterministic In-Network Aggregation (DINA) scheme to improve model training efficiency by enhancing the efficiency of INA utilization in DML. Our key observation is to further increase worker sending rates by reducing gradient packets’ RTT (i.e., realizing packet sub-RTT). Based on this observation, DINA rationally selects the optimal global gradient aggregation switch depending on the switches’ available memory, worker sending rate, and server processing capacity. As a result, DINA reduces the dependence of INA systems on the server, improves worker sending rates, and mitigates network traffic load. We formulate the sub-RTT-INA-based gradient aggregation problem as a mixed-integer nonlinear programming problem. To efficiently solve the problem, we simplify it by transforming the nonlinear constraints into linear constraints and propose a mixed solution that combines randomized rounding and heuristic mechanisms. Simulation results show that DINA can provide determined training by reducing communication time by 12%-17% and network load by 28%-50% compared with existing solutions, thus taking full advantage of INA and realizing a determined INA service.

Parameter Server (PS) is a typical Distributed Machine Learning (DML) enabler and widely used in industry and academia. Existing works propose to apply the emerging In-Network Aggregation (INA) technique to improve model training efficiency. However, existing INA systems may suffer from undetermined model training efficiency and service quality, given that many gradient aggregation processes are still performed by the server under irrational gradient aggregation strategies. In this paper, we propose a Deterministic In-Network Aggregation (DINA) scheme to improve model training efficiency by enhancing the efficiency of INA utilization in DML. Our key observation is to further increase worker sending rates by reducing gradient packets’ RTT. Based on this observation, DINA can rationally select the optimal global gradient aggregation switch depending on the switches’ available memory, worker sending rate, and server processing capacity. Simulation results show that DINA can provide determined training by improving worker sending rates by 16%-87% and network load by 28%-46.8% compared with existing solutions.

Hyperspectral anomaly detection (AD) is important in Earth observation and remote sensing. However, the low spatial resolution of hyperspectral images, insufficient samples and lack of prior information limit the detection accuracy. To solve these problems, in this paper, we propose an auxiliary classifier generative adversarial network model based on a three-dimensional (3D) convolutional neural network named 3D AC-GAN. Firstly, the model is based on a 3D convolutional neural network design, with 3D tensors as samples. The network maintains valuable image spatial spectrum joint features to achieve good detection results. It can also generate sufficient samples to achieve dataset augmentation, solving the overfitting problem in GAN training. Secondly, we train the model with a weakly supervised method. The label of the samples is obtained through the coarse scanning method. Then, the AC-GAN is trained with the bootstrapping method to mitigate the impact of noise labels. The experimental results show that our proposed algorithm outperforms state-of-the-art AD algorithms.

In this paper, we propose an automatic GUI testing approach for Android applications. With the goal of exploring unexecuted event handlers of the application under test as quickly possible, our approach constructs and maintains a dynamic GUI model of the application at run time, which is based on extended non-deterministic labelled transition system that records the weight of transitions between GUI states. Extracting only part of the GUI features of the application under test, the model keeps itself simple enough to avoid state explosion and improves the test efficiency, but provides targeted guidance for testing event generation at the same time. A practical probabilistic search-based event selection algorithm is used to leverage information provided by the model, transform weight of transitions to priority of candidate events, and select the testing event to execute. The algorithm solves the non-deterministic problem introduced by the approximation of the model. Empirical evaluation on several real world applications shows that our approach can achieve high code coverage quickly and detect bugs efficiently.