Real-world data is often unbalanced and exhibits long-tailed distribution over classes. Vanilla classification models trained on imbalanced datasets inherently exhibit bias towards dominant classes. Existing debiasing...
详细信息
Early detection and assessment of polyps play a crucial role in the prevention and treatment of colorectal cancer (CRC). Polyp segmentation provides an effective solution to assist clinicians in accurately locating an...
详细信息
Traditional knowledge distillation adopts one teacher model to instruct the training of a lightweight student model. To improve the performance of knowledge distillation, multi-teacher knowledge distillation utilizing...
详细信息
ISBN:
(数字)9798350359312
ISBN:
(纸本)9798350359329
Traditional knowledge distillation adopts one teacher model to instruct the training of a lightweight student model. To improve the performance of knowledge distillation, multi-teacher knowledge distillation utilizing multi-party knowledge to guide model learning was proposed. However, existing methods mainly use the summation-result of teacher models of a single sample to integrate multi-party knowledge directly, ignoring the relation knowledge among samples. In this paper, we propose a novel multi-teacher knowledge distillation method, which utilizes the data relation knowledge to allocate weights for teachers adaptively. To obtain the data relation knowledge, we design output-based strategy and feature-based strategy, which would help to allocate more weight for the teacher who has better learned the data relation knowledge. Extensive experiments have demonstrated the performance and efficiency of our proposed multi-teacher knowledge distillation method.
Graph convolutional networks (GCNs) are popular for a variety of graph learning tasks. ReRAM-based processing-in-memory (PIM) accelerators are promising to expedite GCN training owing to their in-situ computing capabi...
详细信息
A study is conducted on the LH5T-13.5M electric crane, focusing on the swing characteristics of the load during steel coil lifting operations and the displacement issue of the crane. Initially, the system model is for...
详细信息
This paper presents our error tolerable system for coreference resolution in CoNLL-2011(Pradhan et al., 2011) shared task (closed track). Different from most previous reported work, we detect mention candidates based ...
详细信息
Basecalling is a crucial step in nanopore sequencing as it transforms the raw electrical signal obtained from the nanopore into a readable sequence. The accuracy of basecalling directly affects the quality and reliabi...
Basecalling is a crucial step in nanopore sequencing as it transforms the raw electrical signal obtained from the nanopore into a readable sequence. The accuracy of basecalling directly affects the quality and reliability of the sequenced data. Various algorithms and models are employed to perform accurate basecalling, with advancements in deep learning techniques. However, due to the difficulty of combining feature extraction capability, high parallelism and long sequence modeling capability of existing models, the accuracy and speed of basecalling model are still bottlenecks in data *** this paper, we present TransCaller, an end-to-end accelerated transformer-based nanopore basecaller model. TransCaller comprises a low-parameter electrical signal sampler, a fully transformer-based encoder integrating multiple algorithm-specific modules, and a CTC decoder enhanced with a filter. To further refine and enhance the TransCaller model, we introduce a three-stage pyramid model structure and leverage knowledge distillation techniques to balance the accuracy and speed. Experimental results on 9 datasets demonstrate that TransCaller achieves an average accuracy of 92.95%, surpassing the state-of-the-art Bonito sup model by a margin of 0.65% to 2.34%. Moreover, the distilled TransCallermodel showcases a remarkable 2.97× acceleration in inference speed compared to the original model, effectively catering to the variegated requirements of distinct inference speeds and precision levels across a wide spectrum of scenarios.
Business districts are urban areas that have various functions for gathering people,such as work,consumption,leisure and *** to the dynamic nature of business activities,there exists significant tidal effect on the bo...
详细信息
Business districts are urban areas that have various functions for gathering people,such as work,consumption,leisure and *** to the dynamic nature of business activities,there exists significant tidal effect on the boundary and functionality of business ***,effectively analyzing the tidal patterns of business districts can benefit the economic and social development of a ***,with the implicit and complex nature of business district evolution,it is non-trivial for existing works to support the fine-grained and timely analysis on the tidal effect of business *** this end,we propose a data-driven and multi-dimensional framework for dynamic business district ***,we use the large-scale human trajectory data in urban areas to dynamically detect and forecast the boundary changes of business districts in different time ***,we detect and forecast the functional changes in business *** results on real-world trajectory data clearly demonstrate the effectiveness of our framework on detecting and predicting the boundary and functionality change of business ***,the analysis on practical business districts shows that our method can discover meaningful patterns and provide interesting insights into the dynamics of business *** example,the major functions of business districts will significantly change in different time periods in a day and the rate and magnitude of boundaries varies with the functional distribution of business districts.
With the development of continuous fiber-reinforced composites (CFRCs) 3D printing technology, timely, efficient and accurate detection of fiber path defects is essential for ensuring product quality and performance. ...
详细信息
Public cloud providers embrace multi-tenancy as a strategy to enhance the utilization and efficiency of resources. However, co-located virtual machines (VMs) suffer from qualityof-service (QoS) degradation caused by s...
详细信息
ISBN:
(数字)9798350387117
ISBN:
(纸本)9798350387124
Public cloud providers embrace multi-tenancy as a strategy to enhance the utilization and efficiency of resources. However, co-located virtual machines (VMs) suffer from qualityof-service (QoS) degradation caused by shared resource interference. Existing solutions for predicting QoS degradation often rely on the assumption of online access to application-level information. However, in a production environment, this assumption proves invalid as the VMs are black boxes to the providers. This intrinsic characteristic of the IaaS cloud necessitates the prediction model to generalize to unfamiliar applications and imposes specific criteria on the monitorable *** meet the black-box scenario under Infrastructure as a Service (IaaS) cloudcomputing, we present a novel framework, FEDGE, that can predict interference-aware QoS (IA-QoS) of co-located VMs using only low-level monitorable metrics before migration. Specifically, FEDGE utilizes a stochastic gates layer to select the most informative features from the high-dimensional resource and hardware metrics, which helps to reduce the monitoring overhead. Furthermore, we design a multi-domain MMD-based adversarial denoising autoencoder to regularize the learned hidden representations and prevent over-fitting on the source domains. Next, we employ a multi-layer perceptron (MLP) to accurately predict complex QoS degradation using the learned representations with domain generalization. Experimental results demonstrate that FEDGE outperforms other state-of-the-art methods in terms of both generalizability and effectiveness.
暂无评论