In the field of medical image analysis, precise cancer pathology image segmentation is crucial for assisting clinicians in identifying lesion areas. This paper proposes an innovative dual-branch architecture, DBE-SwinUNet, to overcome the reliance on expert knowledge, inefficiencies, and the limitations of local feature extraction in traditional deep learning models. We introduced depthwise separable convolution (DSConv) to enhance global feature representation and integrated Multi-Scale Dense Skip Connections (MSDSC) to facilitate information flow between the encoder and decoder, improving model stability and segmentation accuracy. Additionally, we incorporated a convolutional neural network enhancement module (DA-DWConv) with a dual attention mechanism, which enhances local feature extraction while effectively controlling model complexity. Experimental results show that on the gastric cancer pathology image (GCPI) dataset, our method achieved a Dice similarity coefficient (DSC) of 90.15 % and a 95th percentile Hausdorff distance (HD95) of 39.25 mm. On the public Synapse dataset, the method reached 80.42% DSC and 19.28 mm HD95, significantly outperforming existing methods.

Recently, the bio-inspired spike camera with continuous motion recording capability has attracted tremendous attention due to its ultra high temporal resolution imaging characteristic. Such imaging feature results in huge data storage and transmission burden compared to that of traditional camera, raising severe challenge and imminent necessity in compression for spike camera captured content. Existing lossy data compression methods could not be applied for compressing spike streams efficiently due to integrate-and-fire characteristic and binarized data structure. Considering the imaging principle and information fidelity of spike cameras, we propose a novel Reconstruction-based Contextual Spike Compression (RCSC) framework, which contains scene reconstruction, contextual image compression and spike generation. To our knowledge, it is the first learning-based model for efficient and robust spike stream compression with informative fidelity. Extensive experimental results show that our model outperforms the state-of-the-art conventional codec VVC intra by 6.14% and surpasses the state-of-the-art learned codec by 2.53% in BD-rate reduction, establishing a strong baseline for spike compression.

An ideal full-reference image quality (FR-IQA) model should exhibit both high separability for images with different quality and compactness for images with the same or indistinguishable quality. However, existing learning-based FR-IQA models that directly compare images in deep-feature space, usually overly emphasize the quality separability, neglecting to maintain the compactness when images are of similar quality. In our work, we identify that the perception bias mainly stems from an inappropriate subspace where images are projected and compared. For this issue, we propose a Debiased Mapping based quality Measure (DMM), leveraging orthonormal bases formed by singular value decomposition (SVD) in the deep features domain. The SVD effectively decomposes the quality variations into singular values and mapping bases, enabling quality inference with more reliable feature difference measures. Extensive experimental results reveal that our proposed measure could mitigate the perception bias effectively and demonstrates excellent quality prediction performance on various IQA datasets.

In recent years, Learned Image Compression (LIC) has undergone rapid evolution. However, it is worthy noting that most prevalent LIC methodologies still rely on uniform Scalar Quantization (SQ) for latent features. This overlooks the untapped potential of contextual information, which could be leveraged to significantly reduce statistical redundancies. Prior researches have explored Vector Quantization (VQ)’s adaptability to diverse data distributions, yet it introduces significant computational complexity into LIC, hindering its practical implementation. Consequently, in this work, we propose the Contextual Sequential Quantization (CSQ) method, which progressively discretizes the latent features of LIC by harnessing content contextual information and image textural priors. Our proposed CSQ signifies progress in LIC by blending the computational efficiency of SQ with a substantial approach towards the adaptability of VQ. We further propose the Center Compensation Module (CCM) based on the proposed CSQ. This module strategically determines adaptive quantization centers, leading to a direct enhancement of reconstruction quality without compromising the bit-rate. Moreover, it is worth noticing that existing LIC approaches face challenges in leveraging hyper side information to effectively enhance transformations, which is attributed to the entanglement of the hyperprior generation module with the main transformations. Consequently, we propose to decouple the hyperprior module from main transformations, and design the Hyperprior-Assisted Transformation (HAT) unit to feed hyperprior back into main transformations. This further improves the coding performance. By integrating all together the proposed CSQ, CCM, and HAT, our proposed Non-uniform quantization-based LIC (NLIC) method attains state-of-the-art rate-distortion (R-D) performance among existing LIC methodologies.

Sparse matrix vector multiplication (SpMV) is a critical performance bottleneck for numerical simulation and artificial intelligence training. The new generation of Sunway supercomputer is the advanced exascale supercomputer in China. The SW26010-Pro many-core processor renders itself as a competitive candidate for its attractive computational power in both numerical simulation and artificial intelligence training. In this paper, we propose a heterogeneous and SPM-aggregated SpMV kernel, specifically designed for the SW26010-Pro many-core processor. To fully exploit the computational power of the SW26010-Pro and balance the load of each core group(CG) during computation, we employ asynchronous computation workflow and propose the SPM-aggregated strategy and vector adaptive mapping algorithm. In addition, we propose the two-level data partition scheme to implement computational load balance. In order to improve memory access efficiency, we directly access memory via DMA controller to replace the discrete memory access. Using several optimizations, we achieve a 77.16x speedup compared to the original implementation. Our experimental results show that the hsSpMV yields up to 3.82× speedups on average compared to the SpMV kernel of the state-of-the-art Sunway math library xMath2.0.

Due to its efficiency and versatility, the preconditioned conjugate gradient algorithm has long been a staple in the realm of iterative linear system solvers. In this paper, we proposed an optimized preconditioned conjugate gradient algorithm tailored for the SW26010-Pro manycore processor, The main work includes: Optimizing data block sizes based on the processor’s storage structure; combining thread-level and data-level parallelism, utilizing manual SIMD to improve efficiency; optimizing the memory access pattern by employing direct memory access and overlapping of computation and communication; employing a shared-memory approach to store long vectors across all cores. Furthermore, we design an accelerated algorithm for reduction operations, to avoid data communication. Experimental results show that the hcaPCG yields up to 28.1× speedups on average compared to the original implementation.

The pseudo random number generators (PRNG) is one type of deterministic functions. The information entropy of the output sequences depends on the entropy of the input seeds. The output sequences can be predicted if attackers could know or control the input seeds of PRNGs. Against that, it is necessary that the input seeds is unpredictable, that is to say, the information entropy of the seeds is high enough. However, if there is no high enough entropy sources in environment, how to generate the seeds of PRNG? In other words, how to increase the entropy of the input seeds? Many approaches for extracting entropy from physical environment have been proposed, which lack of theoretical analysis. The condition of entropy's increasing is given. A model is built to verify the condition based on the functional programming language F*. An example of entropy's increasing is proposed utilizing execution time randomness of arbitrary codes. Then an algorithm is described, which can generate the seed when the entropy value is given.

When computing integral curves and integral surfaces for large-scale unsteady flow fields, a major bottleneck is the widening gap between data access demands and the available bandwidth (both I/O and in-memory). In this work, we explore a novel advection-based scheme to manage flow field data for both efficiency and scalability. The key is to first partition flow field into blocklets (e.g. cells or very fine-grained blocks of cells), and then (pre)fetch and manage blocklets on-demand using a parallel key-value store. The benefits are (1) greatly increasing the scale of local-range analysis (e.g. source-destination queries, streak surface generation) that can fit within any given limit of hardware resources; (2) improving memory and I/O bandwidth-efficiencies as well as the scalability of naive task-parallel particle advection. We demonstrate our method using a prototype system that works on workstation and also in supercomputing environments. Results show significantly reduced I/O overhead compared to accessing raw flow data, and also high scalability on a supercomputer for a variety of applications.

Traffic flow forecasting is an important issue for the application of Intelligent Transportation Systems (ITS). How to improve the traffic flow forecasting precision is a crucial problem. Traffic models in different time sections have great differences. The forecasting precision could be improved if the traffic flow forecasting models were built on different time sections respectively. Traffic flow forecasting usually is real-time and too many forecasting variables will reduce the real-time performance. So the selection of the most informative forecasting variable combination is significant. It can save computation cost and improve forecasting precision. In this paper, information bottleneck theory based on extended entropy is used to partition traffic flow of a day into different time sections. Corresponding to each time section, feature selection based on mutual information is generalized to regression problems and is used to select the most informative variable combination. Selected variables are input to Support Vector Machines (SVM) for traffic flow forecasting. Bayesian inference is used to determine the kernel parameters of SVM. The efficiency of the method is illustrated through analyzing the traffic data of Jinan urban transportation.