Point Cloud Neural Networks (PCNNs) have emerged as a vital tool for latency-sensitive 3D perception applications, such as autonomous driving and AR/VR. However, their inherent computational redundancy—arising from excessive global sampling/search operations and repeated feature updates/aggregations caused by shared neighbors—severely constrains execution efficiency. More critically, conventional redundancy elimination methods usually introduce operations that are highly GPU-unfriendly, resulting in high memory overhead, increased branch divergence, irregular memory access, and serial dependencies, which together pose a significant challenge to PCNN acceleration.

As global demand for efficient network services increases, the limitations of traditional terrestrial wireless networks are becoming more apparent. Space-air-ground integrated networks (SAGIN) have emerged as a promising solution to advance next-generation network infrastructure. However, SAGIN faces significant security challenges—including the lack of a robust security architecture, trust and data reliability issues in multi-hop transmissions, and the need to enhance network performance without compromising security—rendering traditional boundary-based defenses inadequate. Therefore, we propose SECURELINK, a decentralized zero-trust architecture tailored for SAGIN, which replaces traditional perimeter defenses with a ”never trust, always verify” approach. This approach strengthens network security and flexibility through continuous verification and adherence to the principle of least privilege. SECURELINK integrates blockchain technology to establish a multi-layered security verification and data processing scheme, addressing the dynamic and decentralized features of SAGIN. Additionally, we introduce DFRIO, a zero-trust traffic offloading method based on decentralized federated reinforcement learning and blockchain, designed to enhance network performance within maintaining security. Simulation results demonstrate that our solution significantly enhances SAGIN’s defense capability without compromising network stability, outperforming the two baseline schemes by 18.3% and 42.6%, respectively.

Causal inference and root cause analysis play a crucial role in network performance evaluation and optimization by identifying critical parameters and explaining how the configuration parameters affect network key performance indicators (KPIs). Traditional performance evaluation methods can evaluate KPIs based on configuration parameters, but they are unable to explain how configuration parameters affect KPIs. Moreover, static causal discovery and inference methods are not directly applicable to dynamic networks. To address these challenges, we propose a self-iterative performance optimization method based on causal inference and interpretable network evaluation (CiiNet). CiiNet constructs causal graphs through change-point detection and hierarchical incremental causal discovery. Then, CiiNet introduces causal inference for critical parameter analysis (CPA). Using intervention analysis and regression-based parameter learning, CiiNet infers and evaluates the impact of critical parameters on KPIs. Based on the interpretable evaluation, CiiNet can further self-iteratively optimize the critical parameters for optimal network performance and dynamically obtain the optimal configurations. Our extensive experiments show that CiiNet outperforms other baseline methods regarding causal discovery, network performance evaluation, and CPA.

Fault diagnosis in microservice systems requires high availability, driving research towards multimodal learning that leverages heterogeneous monitoring data, including logs, metrics, and traces. These data inherently contain time series and data streams across various modalities. However, existing frameworks fail to exploit temporal dependencies in these dynamic streams effectively. First, existing methods rely too much on original time-domain metrics, which makes it difficult for them to capture periodic patterns and sudden events. Second, most language-bound integrations treat timestamps only as sequential labels or numerical inputs, which ignores the rich contextual information within timestamps. As a result, existing models struggle to comprehend the temporal sequences and data characteristics associated with faults in multimodal data. In this work, we introduce FD-TE Diagnosis, a framework for microservice fault diagnosis that addresses these challenges by applying frequency domain feature analysis and time encoding. To enhance the detection of abnormal events, we integrate a frequency-domain approach that utilizes the Fast Fourier Transform (FFT) to extract robust features from metric data. Next, we encode the timestamps of all events using a dedicated time encoding layer. These temporal representations are then incorporated to strengthen event embedding for fault inference. Experimental results demonstrate that our method enhances the sensitivity of multimodal models in interpreting temporal features, resulting in improved diagnostic outcomes.

Anomaly detection in multivariate time series data is critical across a variety of real-life applications. The predominant anomaly detection techniques currently rely on reconstruction-based methods. However, these methods often overfit the abnormal pattern and fail to diagnose the anomaly. Although some studies have attempted to prevent the incorrect fitting of anomalous data by enabling models to learn the trend of data variations, they fail to account for the dynamic nature of data distribution. This oversight can lead to the erroneous reconstruction of anomalies that do not exist. To address these challenges, we propose RTdetector, a Transformer-based time series anomaly detection model leveraging reconstruction trends. RTdetector employs a novel global attention mechanism based on reconstruction trends to learn distinguishable attention from the original sequence, thereby preserving the global trend information intrinsic to the time series. Additionally, it incorporates a self-conditioning transformer, based on reconstruction trend enhancement to achieve superior predictive performance. Extensive experiments on four datasets demonstrate that RTdetector achieves state-of-the-art results in multivariate time series data anomaly detection. Our code is available at https://github.com/CSUFUNLAB/RTdetector.

The exponential growth of the Internet of Things (IoT) necessitates the deployment of large-scale models capable of processing the complex and diverse data generated by IoT devices. However, the substantial memory requirements of these models pose significant challenges, especially in scenarios where rapid decision-making and low-latency responses are critical. To address these challenges, we propose three innovative strategies for optimizing large model usage in IoT environments. The first strategy is an adaptive loading scheme, which enables dynamic loading of individual model experts. The second strategy involves an expert-by-expert loading approach, further enhancing the ability to load experts as needed, which optimizes memory usage and accelerates computations. The third strategy employs an interlayer expert reuse mechanism, facilitating the efficient reuse of experts across different layers, thus enhancing response rates without compromising model accuracy. Importantly, these strategies can be directly applied to Mixture of Experts (MoE) large language models without requiring additional training, thereby providing a seamless and efficient solution for leveraging these models in memory-constrained, high-performance IoT environments.

The Electroencephalogram (EEG)-based Brain-Computer Interface (BCI) is important for Internet of Things (IoT) applications. EEG data can be used to control IoT devices for applications such as smart home automation or healthcare monitoring. EEG-based BCI systems are crucial for recognizing human brain thoughts and analyzing neurological diseases, enabling thought visualization, and improving accessibility for people with disabilities. With the rapid development of machine learning, including deep learning technologies, wearable BCI devices, and hybrid BCI research have advanced significantly, showcasing their remarkable advantages. Researchers have conducted extensive experiments to improve the accuracy of the system. This paper provides a comprehensive review of BCI based on EEG, highlighting the fundamental principles of EEG signals, common acquisition devices, feature extraction techniques, and classification models, with a particular focus on the latest advances in deep learning. We also summarize available datasets and discuss the latest applications of EEG-based BCI in human-computer interaction and neurological diseases. Finally, we highlight the main findings and explore future directions, offering researchers deeper insight to foster further progress in this field.

2-hop labeling has been widely utilized to accelerate the efficiency of online shortest distance queries. Given the nature of frequent changes in real-world graphs, the efficient maintenance of 2-hop labeling index has been extensively studied recently. However, existing methods cannot efficiently process large-scale graphs due to their high time and memory costs, and most of them process large batches of updates sequentially, significantly decreasing efficiency. In this paper, we propose a novel algorithm for maintaining the 2-hop labeling index in a parallel manner, called M2HL , which can efficiently handle both edge insertions and deletions. Moreover, we theoretically prove that M2HL maintains both correctness and minimality for the updated 2-hop labeling index. Our experiments on ten large-scale graphs demonstrate that M2HL outperforms the state-of-the-art 2-hop labeling maintenance methods by up to four orders of magnitude in speed while maintaining correctness and minimality, as well as exhibiting strong scalability and low memory usage.

While vision transformers (ViTs) have continued to achieve new milestones in computer vision, their complicated network architectures with high computation and memory costs have hindered their deployment on resource-limited edge devices. Some customized accelerators have been proposed to accelerate the execution of ViTs, achieving improved performance with reduced energy consumption. However, these approaches utilize flattened attention mechanisms and ignore the inherent hierarchical visual semantics in images. In this work, we conduct a thorough analysis of hierarchical visual semantics in real-world images, revealing opportunities and challenges of leveraging visual semantics to accelerate ViTs. We propose ReViT, a systematic algorithm and architecture co-design approach, which aims to exploit the visual semantics to accelerate ViTs. Our proposed algorithm can leverage the same semantic class with strong feature similarity to reduce computation and communication in a differential attention mechanism, and support the semantic-aware attention efficiently. A novel dedicated architecture is designed to support the proposed algorithm and translate it into performance improvements. Moreover, we propose an efficient execution dataflow to alleviate workload imbalance and maximize hardware utilization. ReViT opens new directions for accelerating ViTs by exploring the underlying visual semantics of images. ReViT gains an average of 2.3$\boldsymbol{\times}$speedup and 3.6$\boldsymbol{\times}$energy efficiency over state-of-the-art ViT accelerators.

Point cloud neural networks are gaining increasing attention in emerging 3-D computer vision applications, such as autonomous driving, robotics, and virtual reality. Many customized accelerators for 3-D point clouds have been developed to pursue superior time and energy efficiencies. In this work, we reveal that spatially adjacent points in a 3-D point cloud show similar feature values and relationships, implying substantial redundant computations and memory accesses, while which have been previously ignored. To reduce such redundancies, we propose SimDiff, an algorithm-accelerator co-design framework that boosts 3-D point cloud processing by cleverly leveraging spatial similarity toward excellent speedup and energy efficiency. On the algorithm side, we design a novel similarity-aware differential point cloud neural network (dubbed SD-PCNet). Differing from the standard flow of mainstream point cloud networks, it abstracts a brand-new execution flow for point cloud processing by utilizing spatial similarity among points and dynamic differential execution. On the accelerator side, we propose SD-PCAcc, a supporting accelerator to convert algorithm-level redundancy reductions into performance enhancements. On the deployment side, we propose efficient strategies for network-to-accelerator mapping and scheduling, high-bandwidth memory (HBM) channel allocation, and core component reconfiguration, facilitating the proposed methodologies into practical implementation. Extensive evaluation results show that, with preserved accuracy, our SimDiff gains an average of$3.2\times $speedup and$3.1\times $energy efficiency compared to the state-of-the-art competitors.

Highly dynamic heterogeneous networks (HDHNs), characterized by high node mobility and heterogeneity, frequently experience complex and recurrent network faults. Conventional centralized fault diagnosis methods demand real-time collection of extensive network-wide data, while distributed approaches often exhibit limited fault detection capabilities. Additionally, machine learning-based fault diagnosis methods are challenged by the scarcity of labeled fault samples required for training. To address these limitations, this study proposes a semi-distributed network fault diagnosis architecture based on a digital twin network (DTN). The proposed architecture facilitates the extraction of a comprehensive labeled fault dataset that closely replicates real-world network conditions. Using this dataset, we perform centralized training of an enhanced anomaly detection model, FTS-LSTM, to infer fault types at the node level. To overcome the drawbacks of both centralized and distributed approaches, we further introduce a semi-distributed fault diagnosis algorithm (SDFD) that integrates fault types and severity levels identified by nodes to infer overall network faults. The proposed fault diagnosis scheme is validated on a semi-physical DTN simulation platform, demonstrating its effectiveness in realistic scenarios.

Packet-level load balancing has shown its massive potential for long in utilizing super high bisection bandwidth of data center network (DCN). This kind of potential, however, has still not been completely transformed into huge performance enhancement of data transmission. The fundamental reason is that packet-level load balancing can fully utilize the parallel paths of underlying physical network, but suffer from the problem of packet disordering transmission, which greatly impairs the flow-level transmission performance of DCN. This paper explores the root cause of performance impairment generated by packet disordering transmission, and proposes$R^{3}$, a solution focusing on “recognizably releasing redundant acknowledgements” as a building block for data center packet-level load balancer. In$R^{3}$'s heart, the source leaf switch perceives the global packet loss information and selectively intercepts the redundant acknowledgement packets, thus avoiding the TCP-driven end-host from experiencing frequent window reductions and unnecessary packet retransmissions. Experimental results of numerous simulation tests and real implementations show that, after integrating$R^{3}$into the representative data center packet-level load balancing schemes, the transmission performances of both delay-sensitive and throughput-oriented data center flows are significantly improved. Furthermore,$R^{3}$is merely implemented by switch, leaving the end hosts and the deployed load balancing scheme totally unchanged.

In the domain of medical diagnostics, precise identification of various skin and oral diseases is vital for effective patient care. In particular, Mpox is a potentially dangerous viral disease with zoonotic origins, capable of human-to-human transmission, underscoring the urgency of precise diagnostic methods for timely intervention. This paper introduces a novel approach named the Choquet Fuzzy Integral-based Ensemble (CFI-Net) for accurate classification of skin diseases, with a specific emphasis on detecting Mpox, foot ulcers, and various mouth and oral diseases. Our methodology begins with Transfer Learning, enhancing the classification capabilities of base classifiers (DenseNet169, MobileNetV1 and DenseNet201) by incorporating additional layers. Subsequently, we aggregate the prediction scores from each base classifier using the Choquet fuzzy integral (CFI) to derive the final predicted labels, thus ensuring dynamic and robust predictions. Fuzzy measures, a crucial component of this fuzzy integral-based ensemble method, are typically determined through manual experimentation in previous approaches. However, in our study, we have tackled the challenge of manual tuning by employing meta-heuristic optimization algorithm to precisely configure the fuzzy measures for optimal performance. A rigorous evaluation is conducted on four publicly available datasets, encompassing two Mpox datasets, a foot ulcer dataset, and a mouth and oral disease dataset. The experiments reveal the remarkable effectiveness of CFI-Net in significantly improving disease classification accuracy. Additionally, we employ Grad-CAM analysis to provide insights into the decision-making processes of our models. Our findings underscore the exceptional performance of CFI-Net, achieving accuracy rates of 98.06% and 94.81% for Mpox detection, 99.06% for foot ulcer detection, and an impressive 99.61% for mouth and oral disease classification. This research not only contributes to the advancement of disease diagnosis but also demonstrates the effectiveness of ensemble learning techniques coupled with fuzzy integral-based fusion in enhancing diagnostic accuracy.

Point clouds play a crucial role in various fields such as robotics, 3D modeling, and autonomous driving. DGCNN, as a representative work in this domain, has exhibited superior performance compared to classic models like PointNet++. Such models often display significant computational demands and encounter notable performance bottlenecks, so how to accelerate their calculations has become a noteworthy issue. Nevertheless, due to the inherent disparities between point cloud tasks and conventional convolution tasks, existing algorithms and architectural paradigms designed for convolution frequently prove to be ill-suited for point clouds, such as Tensor Core (TC). There have been studies focused on reducing DGCNN model complexity to improve computational efficiency, but research specifically tailored to TC acceleration remains relatively limited. Therefore, we present an encapsulation module that bridges the computation of model in Torch level with the components of the CUDA backend, introducing TC into DGCNN. It dynamically adjusts parameters and computation flow based on input data and transforms batch computations into streaming computations. Additionally, we propose a new computational organization approach, which involves the restructuring of two critical computation steps within DGCNN: the Get Graph Feature operation and the Conv operation. It significantly enhances the computational speed of these two operations while reducing memory redundancy of tensor data involved in the intermediate processes. In the experiment, our modifications achieve an average acceleration of 1.80X for the Get Graph Feature operation and 1.46X for the Conv operation, all while not compromising result accuracy.

Deep learning on point clouds has attracted increasing attention for various emerging 3D computer vision applications, such as autonomous driving, robotics, and virtual reality. These applications interact with people in real-time on edge devices and thus require low latency and low energy. To accelerate the execution of deep neural networks (DNNs) on point clouds, some customized accelerators have been proposed, which achieved a significantly higher performance with reduced energy consumption than GPUs and existing DNN accelerators.

With the explosive growth of the Internet of Things (IoT), an increasing amount of sensor data generated by soft real-time IoT applications has been moved to data centers for storage and data analysis. Large amounts of these data are required to be processed within a given deadline to ensure application performance. Therefore, meeting the transmission deadlines of data flows for soft real-time applications has always been crucial yet challenging to current data centers. Recent progress has demonstrated that adopting parallel data transmission over multipath data center network combining with effective load balancing can achieve a high bisection network bandwidth, thus speeding up the network transfer of data flows. Nevertheless, the deadline miss ratios (DMRs) of these flows are not lowered as expected since the existing load balancing schemes are naturally agnostic to the deadline requirement. They are either unable to reroute traffic flexibly or aimlessly reroute these deadline-restrained flows, regardless of their urgent levels and path conditions. To address these inefficiencies, we propose a deadline-aware load-balancing scheme, namely, DLB, which perceives the deadline requirements and helps the urgent flows to timely switch to those faster transmission paths to complete quickly. Specifically, DLB computes the urgent level for each flow in real time to judge if the switch needs to make proactive rerouting. When a flow is nonurgent, DLB does not proactively change its transmission path, leaving more available paths to those flows with higher urgent levels. When a flow becomes extremely urgent, it immediately switches to those light-loaded paths to finish its data transmission before its deadline as far as possible. Experimental results of NS2 simulations and real testbed implementations show that DLB reduces the DMRs by up to 50% compared to the state-of-the-art data center load-balancing schemes, while only induces trivial overhead during deployment.

Digital twin technology has recently gathered pace in engineering communities as it allows for the convergence of the real structure and its digital counterpart. 3D point cloud data is a more effective way to describe the real world and to reconstruct the digital counterpart than the conventional 2D images or 360-degree images. Large-scale, e.g., city-scale digital twins, typically collect point cloud data via internet-of-things (IoT) devices and transmit it over wireless networks. However, the existing wireless transmission technology can not carry real-time point cloud transmission for digital twin reconstruction due to mass data volume, high processing overheads, and low delay-tolerance. We propose a novel artificial intelligence (AI) powered end-to-end framework, termed AIRec, for efficient digital twin communication from point cloud compression, wireless channel coding, and digital twin reconstruction. AIRec adopts the encoder-decoder architecture. In the encoder, a novel importance-aware pooling scheme is designed to adaptively select important points with learnable thresholds to reduce the transmission volume. We also design a novel noise-aware joint source and channel coding is proposed to adaptively adjust the transmission strategy based on SNR and map the features to error-resilient channel symbols for wireless transmission to achieve a good tradeoff between the transmission rate and reconstruction quality. The decoder can accurately reconstruct the digital twins from the received symbols. Extensive experiments of typical datasets and comparison with baselines show that we achieve a good reconstruction quality under$24\times $compression ratio.

The cloud-assisted industrial Internet of Things (IIoT) architecture can sustain highly available computation and massive storage services for modern industrial systems. When data owners store IIoT data to remote cloud platforms, the data security will face tough challenges. Cryptographic technologies endow an ability to guarantee data confidentiality. However, traditional encryption techniques make data access control and data searching malfunctioning. Recently emerging attribute-based keyword search (ABKS) primitive achieves fine-grained access control and effective data searching over ciphertexts. However, existing ABKS schemes only consider single data owner scenarios and may be an inappropriate choice for IIoT applications, where there exists multiple data owners for an integrated industrial system. Directly extending state-of-the-art single owner schemes to ones for multiowner environment will impose a complicated key management issue. We present an ABKS scheme for multiowners in the cloud-assisted IIoT architecture. By designing a novel master key generation and private key aggregation mechanism with desired communication overheads, our scheme eliminates the complex key management issue in the multiowner model. Formal security proof demonstrates that our scheme is secure against the cloud server. Experimental evaluations also demonstrate its correctness and practicality.

Analysis of high dimensional biomedical data such as microarray gene expression data and mass spectrometry images, is crucial to provide better medical services including cancer subtyping, protein homology detection, etc. Clustering is a fundamental cognitive task which aims to group unlabeled data into multiple clusters based on their intrinsic similarities. However, for most clustering methods, including the most widely used K-means algorithm, all features of the high dimensional data are considered equally in relevance, which distorts the performance when clustering high-dimensional data where there exist many redundant variables and correlated variables. In this paper, we aim at addressing the problem of the high dimensional bioinformatics data clustering and propose a new correlation induced clustering, CoIn, to capture complex correlations among high dimensional data and guarantee the correlation consistency within each cluster. We evaluate the proposed method on a high dimensional mass spectrometry dataset of liver cancer tumor to explore the metabolic differences on tissues and discover the intra-tumor heterogeneity (ITH). By comparing the results of baselines and ours, it has been found that our method produces more explainable and understandable results for clinical analysis, which demonstrates the proposed clustering paradigm has the potential with application to knowledge discovery in high dimensional bioinformatics data.

Point cloud data, a flexible 3D object representation, is critical for various applications such as autonomous driving, robotics and remote sensing. Despite the recent success of deep neural networks (DNNs) on supervised point cloud analysis tasks, they still rely on tedious manual annotation of point clouds and cannot make predictions for new classes. Unlike few-shot learning for 2D images with the advantages of large-scale datasets and high-quality deep pre-trained models like ResNet, for 3D few-shot learning, obtaining discriminative representations of unseen classes with high intra-class similarity and inter-class difference is very challenging. To address this issue, this work proposes a novel cascade graph neural network for few-shot learning on point clouds, termed as CGNN, in which two cascade GNNs are adopted to extract the intra-object topological information and learn the inter-object relations respectively. To further increase the discriminability of point cloud features, we first design a novel discriminative edge label to model the intra-class similarity and inter-class dissimilarity based on channel-wise feature variance and class consistency. Second, we propose a novel few-shot circle loss which classifies the nodes into two subsets, i.e., support to support pairs and support to query pairs, and optimizes the pair-wise similarity on two subsets independently. Extensive experiments on benchmark CAD and real LiDAR point cloud datasets have demonstrated that CGNN improves accuracy by 5.98% over the state-of-the-art GNN-based few-shot classification methods.

Both accurate and fast mobile recommendation systems based on click behaviors analysis are crucial in e-business. Deep learning has achieved state-of-the-art accuracy and the traditional wisdom often hosts these computation-intensive models in powerful cloud centers. However, the cloud-only approaches put significant computational pressure on cloud servers and increase the latency in heavy-load scenarios. Moreover, existing work often adopts RNN structures to model behaviors that suffer from low processing speed for under-utilization of parallel devices such as GPUs. In this work, we propose an efficient internet behavior-based recommendation framework with edge-cloud collaboration on deep CNNs (CoRec) to improve both the accuracy and speed for mobile recommendation. A novel convolutional interest network (CIN) improves the accuracy by modeling the long- and short-term interests and accelerates the prediction through parallel-friendly convolutions. To further improve the serving throughput and latency, a novel device-cloud collaboration strategy reduces workloads by pre-computing and caching long-term interests in the cloud offline and real-time computation of short-term interests in devices. Extensive experiments on real-world datasets show that CoRec significantly outperforms the state-of-the-art methods in accuracy and has achieved at least an order of magnitude improvement in latency and throughput compared to cloud-only RNN-based approaches for long behaviors.

The attribute-based keyword search (ABKS), which simultaneously achieves searching and fine-grained access control over encrypted data, is frequently applied in cloud computing environments characterized by data storage and sharing. Recently, inspired by attribute-based encryption (ABE) and searchable encryption (SE) primitives, several ABKS schemes have been presented. However, almost all existing ABKS schemes actually only provide an attribute-based keyword equality match function and do not have a structural index to support practical search efficiency and dynamic data updates in real-world applications. To the best of our knowledge, this study is the first to realize an attribute-based keyword search construction supporting numerical comparison expressions with the practical search efficient and dynamic data update capacity (ABKS-NICEST), based on our proposed attribute-based keyword secure search scheme supporting numerical comparison expressions (ABKS-NICE) and anexclusive OR-chain-based inverted index structure. To the best of our knowledge, ABKS-NICEST is the first attributed-based keyword search scheme with practical search efficiency and dynamic data update capacity. In addition, numerical values are an important and common attribute, so providing comparison expressions among numerical values can greatly enhance the expressivity of access policy. Therefore, we use the prefix membership verification technique to design a method to support any numeric comparison expression in a flexible and uniform manner. Through theoretical and experimental evaluations, we determine that ABKS-NICEST is the most efficient ABKS scheme.

Graph neural networks (GNNs), which extend conventional deep learning technologies to process graph-structured data, have shown its powerful graph representation learning ability. Existing typical GNNs utilize neighborhood message passing mechanism based on neural networks that updates target vertex representations by aggregating feature messages from neighboring source vertices. To accelerate the computations of GNNs, some customized accelerators, which follow the neighborhood aggregation computation pattern for each vertex, have been proposed. Through analysis, we observe that a naive implementation of the neighborhood aggregation results in redundant computations and communications.In this paper, we propose a novel redundancy-eliminated GNN accelerator, shortly termed as ReGNN. ReGNN is supported by an algorithm and architecture co-design. We first propose a dynamic redundancy-eliminated neighborhood message passing algorithm for GNNs. Then a novel architecture is designed to support the proposed algorithm and transform the redundancy elimination into performance improvement. ReGNN is also a configurable pipelined architecture that can be configured to support different GNN variants. In terms of the same computations, ReGNN provides the same accuracy as traditional GNNs. To the best of our knowledge, ReGNN is the first accelerator that can eliminate computation redundancy in GNNs. Our proposed ReGNN system gains an average of 9.1× speedup and 8.9× energy efficiency over state-of-the-art GNN accelerators.

Point cloud, an efficient 3D object representation, plays an indispensable role in autonomous driving technologies, such as object avoidance, localization, and map building. The analysis of point clouds (e.g., 3D segmentation) is essential to exploit the informative value of point clouds for such applications. The main challenge remains to effectively and completely extract high-level point cloud feature representations. To this end, we present a novel multi-task Y-shaped graph neural network to explore 3D point clouds, referred to as MTYGNN. By extending the conventional U-Net, MTYGNN contains two main branches to simultaneously perform classification and segmentation tasks in point clouds. Meanwhile, the classification prediction is fused together with the semantic features as the scene context to make the segmentation task more accurate. Furthermore, we consider the homoscedastic uncertainty of each task to calculate the weights of multiple loss functions to ensure that tasks do not negatively interfere with each other. The proposed MTYGNN is evaluated on popular point cloud datasets in traffic scenarios. Experimental results demonstrate that our framework outperforms the state-of-the-art baseline methods.

Time-series forecasting is an important problem across a wide range of domains. Designing accurate and prompt forecasting algorithms is a non-trivial task, as temporal data that arise in real applications often involve both non-linear dynamics and linear dependencies, and always have some mixtures of sequential and periodic patterns, such as daily, weekly repetitions, and so on. At this point, however, most recent deep models often use Recurrent Neural Networks (RNNs) to capture these temporal patterns, which is hard to parallelize and not fast enough for real-world applications especially when a huge amount of user requests are coming. Recently, CNNs have demonstrated significant advantages for sequence modeling tasks over the de-facto RNNs, while providing high computational efficiency due to the inherent parallelism. In this work, we propose HyDCNN, a novel hybrid framework based on fully Dilated CNN for time-series forecasting tasks. The core component in HyDCNN is a proposed hybrid module, in which our proposed position-aware dilated CNNs are utilized to capture the sequential non-linear dynamics and an autoregressive model is leveraged to capture the sequential linear dependencies. To further capture the periodic temporal patterns, a novel hop scheme is introduced in the hybrid module. HyDCNN is then composed of multiple hybrid modules to capture the sequential and periodic patterns. Each of these hybrid modules targets on either the sequential pattern or one kind of periodic patterns. Extensive experiments on five real-world datasets have shown that the proposed HyDCNN is better compared with state-of-the-art baselines and is at least 200% better than RNN baselines. The datasets and source code will be published in Github to facilitate more future work.

Recently, graph neural networks (GNNs) have achieved great success for graph representation learning tasks. Enlightened by the fact that numerous message passing redundancies exist in GNNs, we propose DyGNN, which speeds up GNNs by reducing redundancies. DyGNN is supported by an algorithm and architecture co-design. The proposed algorithm can dynamically prune vertices and edges during execution without accuracy loss. An architecture is designed to support dynamic pruning and transform it into performance improvement. DyGNN opens new directions for accelerating GNNs by pruning vertices and edges. DyGNN gains average $2\times$ speedup with accuracy improvement of 4% compared with state-of-the-art GNN accelerators.

Recent trend focuses on using heterogeneous graph of things (HGoT) to represent things and their relations in the Internet of Things, thereby facilitating the applying of advanced learning frameworks, i.e., deep learning (DL). Nevertheless, this is a challenging task since the existing DL models are hard to accurately express the complex semantics and attributes for those heterogeneous nodes and links in HGoT. To address this issue, we develop attention-aware encoder-decoder graph neural networks for HGoT, termed as HGAED. Specifically, we utilize the attention-based separate-and-merge method to improve the accuracy, and leverage the encoder-decoder architecture for implementation. In the heart of HGAED, the separate-and-merge processes can be encapsulated into encoding and decoding blocks. Then, blocks are stacked for constructing an encoder-decoder architecture to jointly and hierarchically fuse heterogeneous structures and contents of nodes. Extensive experiments on three real-world datasets demonstrate the superior performance of HGAED over state-of-the-art baselines.