To better characterize the properties of surface-initiated polymers, simultaneous bulk-and surface-initiated polymerizations are usually carried out by assuming that the properties of the surface-initiated polymers re...
详细信息
To better characterize the properties of surface-initiated polymers, simultaneous bulk-and surface-initiated polymerizations are usually carried out by assuming that the properties of the surface-initiated polymers resemble those of the bulk-initiated polymers. Through a Monte Carlo simulation using a heterogeneous stochastic reaction model, it was discovered that the bulk-initiated polymers exhibit a higher molecular weight and a lower dispersity than the corresponding surface-initiated polymers, which indicates that the equivalent assumption is invalid. Furthermore, the molecular weight distributions of the two types of polymers are also different, suggesting different polymerization mechanisms. The results can be simply explained by the heterogeneous distributions of reactants in the system. This study is helpful to better understand surface-initiated polymerization.
With the continuous decrease in the critical dimensions of integrated circuits, mask optimization has becomethe main challenge in VLSI design. In recent years, thriving machine learning has been gradually introduced i...
详细信息
With the continuous decrease in the critical dimensions of integrated circuits, mask optimization has becomethe main challenge in VLSI design. In recent years, thriving machine learning has been gradually introduced in the field ofoptical proximity correction (OPC). Currently, advanced learning-based frameworks have been limited by low mask printability or large computational overhead. To address these limitations, this paper proposes a learning-based frameworknamed SegNet-OPC, which can generate optimized masks from the target layout at shorter training and turnaround timewith higher mask printability. The proposed framework consists of a backbone network and loss terms suitable for maskoptimization tasks, followed by a fine-tuning network. The framework yields remarkable improvements over conventionalmethods, delivering significantly faster turnaround time and superior mask printability and manufacturability. With just1.25 hours of training, the framework achieves comparable mask complexity while surpassing the state-of-the-art methods,achieving a minimum 3% enhancement in mask printability and an impressive 16.7% improvement in mask manufacturability.
Internet of Things (IoT) is an evolving paradigm for building smart cross-industry. The data gathered from IoT devices may have anomalies or other errors for various reasons, such as malicious activities or sensor fai...
详细信息
Cloud computing has emerged as a promising mode for storaging vast quantities of big data, which is vulnerable to potential security threats, making it urgent to ensure data confidentiality and integrity auditing. In ...
详细信息
MXene is a promising energy storage material for miniaturized microbatteries and microsupercapacitors(MSCs).Despite its superior electrochemical performance,only a few studies have reported MXene-based ultrahigh-rate(...
详细信息
MXene is a promising energy storage material for miniaturized microbatteries and microsupercapacitors(MSCs).Despite its superior electrochemical performance,only a few studies have reported MXene-based ultrahigh-rate(>1000 mV s^(−1))on-paper MSCs,mainly due to the reduced electrical conductance of MXene films deposited on ***,ultrahigh-rate metal-free on-paper MSCs based on heterogeneous MXene/poly(3,4-ethylenedioxythiophene)-poly(styrenesulfonate)(PEDOT:PSS)-stack electrodes are fabricated through the combination of direct ink writing and femtosecond laser *** a footprint area of only 20 mm^(2),the on-paper MSCs exhibit excellent high-rate capacitive behavior with an areal capacitance of 5.7 mF cm^(−2)and long cycle life(>95%capacitance retention after 10,000 cycles)at a high scan rate of 1000 mV s^(−1),outperforming most of the present on-paper ***,the heterogeneous MXene/PEDOT:PSS electrodes can interconnect individual MSCs into metal-free on-paper MSC arrays,which can also be simultaneously charged/discharged at 1000 mV s^(−1),showing scalable capacitive *** heterogeneous MXene/PEDOT:PSS stacks are a promising electrode structure for on-paper MSCs to serve as ultrafast miniaturized energy storage components for emerging paper electronics.
Vehicular data misuse may lead to traffic accidents and even loss of life,so it is crucial to achieve secure vehicular data *** paper focuses on secure vehicular data communications in the Named Data Networking(NDN).I...
详细信息
Vehicular data misuse may lead to traffic accidents and even loss of life,so it is crucial to achieve secure vehicular data *** paper focuses on secure vehicular data communications in the Named Data Networking(NDN).In NDN,names,provider IDs and data are transmitted in plaintext,which exposes vehicular data to security threats and leads to considerable data communication costs and failure *** paper proposes a Secure vehicular Data Communication(SDC)approach in NDN to supress data communication costs and failure *** constructs a vehicular backbone to reduce the number of authenticated nodes involved in reverse *** the ciphtertext of the name and data is included in the signed Interest and Data and transmitted along the backbone,so the secure data communications are *** is evaluated,and the data results demonstrate that SCD achieves the above objectives.
With the great development of Multi-Target Tracking(MTT)technologies,many MTT algorithms have been proposed with their own advantages and *** to the fact that requirements to MTT algorithms vary from the application s...
详细信息
With the great development of Multi-Target Tracking(MTT)technologies,many MTT algorithms have been proposed with their own advantages and *** to the fact that requirements to MTT algorithms vary from the application scenarios,performance evaluation is significant to select an appropriate MTT algorithm for the specific application *** this paper,we propose a performance evaluation method on the sets of trajectories with temporal dimension specifics to compare the estimated trajectories with the true *** proposed method evaluates the estimate results of an MTT algorithm in terms of tracking accuracy,continuity and ***,its computation is based on a multi-dimensional assignment problem,which is formulated as a computable form using linear *** enhance the influence of recent estimated states of the trajectories in the evaluation,an attention function is used to reweight the trajectory errors at different time ***,simulation results show that the proposed performance evaluation method is able to evaluate many aspects of the MTT *** evaluations are worthy for selecting suitable MTT algorithms in different application scenarios.
Video question answering(VideoQA) is a challenging yet important task that requires a joint understanding of low-level video content and high-level textual semantics. Despite the promising progress of existing efforts...
详细信息
Video question answering(VideoQA) is a challenging yet important task that requires a joint understanding of low-level video content and high-level textual semantics. Despite the promising progress of existing efforts, recent studies revealed that current VideoQA models mostly tend to over-rely on the superficial correlations rooted in the dataset bias while overlooking the key video content, thus leading to unreliable results. Effectively understanding and modeling the temporal and semantic characteristics of a given video for robust VideoQA is crucial but, to our knowledge, has not been well investigated. To fill the research gap, we propose a robust VideoQA framework that can effectively model the cross-modality fusion and enforce the model to focus on the temporal and global content of videos when making a QA decision instead of exploiting the shortcuts in datasets. Specifically, we design a self-supervised contrastive learning objective to contrast the positive and negative pairs of multimodal input, where the fused representation of the original multimodal input is enforced to be closer to that of the intervened input based on video perturbation. We expect the fused representation to focus more on the global context of videos rather than some static keyframes. Moreover, we introduce an effective temporal order regularization to enforce the inherent sequential structure of videos for video representation. We also design a Kullback-Leibler divergence-based perturbation invariance regularization of the predicted answer distribution to improve the robustness of the model against temporal content perturbation of videos. Our method is model-agnostic and can be easily compatible with various VideoQA backbones. Extensive experimental results and analyses on several public datasets show the advantage of our method over the state-of-the-art methods in terms of both accuracy and robustness.
Large-scale GPU clusters are widely used to speed up both latency-critical(online) and besteffort(offline) deep learning(DL) workloads. However, similar to the common practice, the DL clusters at ByteDance dedicate ea...
详细信息
Large-scale GPU clusters are widely used to speed up both latency-critical(online) and besteffort(offline) deep learning(DL) workloads. However, similar to the common practice, the DL clusters at ByteDance dedicate each GPU to one workload or share workloads in time dimension, leading to very low GPU resource utilization. Existing techniques like NVIDIA MPS provide an opportunity to share multiple workloads in space on widely-deployed NVIDIA GPUs, but it cannot guarantee the performance of online workloads. We present MuxFlow, the first production system that can scale over massive GPUs to support highly efficient space-sharing for DL workloads. MuxFlow introduces a two-level protection mechanism for both memory and computation to guarantee the performance of online workloads. MuxFlow leverages dynamic streaming multiprocessor(SM) allocation to improve the efficiency of offline workloads. Based on our practical error analysis, we design a mixed error-handling mechanism to improve system *** has been deployed at ByteDance on more than 18000 GPUs. The deployment results indicate that MuxFlow substantially improves the GPU utilization from 26% to 76%, SM activity from 16% to 33%, and GPU memory usage from 42% to 48%.
A core challenge in applying deep reinforcement learning (DRL) to real-world tasks is the sparse reward problem, and shaping reward has been one effective method to solve it. However, due to the enormous state space a...
详细信息
暂无评论