Large Language Models (LLMs) have shown promising performance in text-to-SQL, which involves translating natural language questions into SQL queries. However, current text-to-SQL LLMs are computationally expensive and...
详细信息
Vortices are whirling disturbances,commonly found in nature,ranging from tremendously small scales in Bose-Einstein condensations to cosmologically colossal scales in spiral *** optical vortex,generally associated wit...
详细信息
Vortices are whirling disturbances,commonly found in nature,ranging from tremendously small scales in Bose-Einstein condensations to cosmologically colossal scales in spiral *** optical vortex,generally associated with a spiral phase,can carry orbital angular momentum(OAM).The optical OAM can either be in the longitudinal direction if the spiral phase twists in the spatial domain or in the transverse direction if the phase rotates in the spatiotemporal *** this article,we demonstrate the intersection of spatiotemporal vortices and spatial vortices in a wave *** a result of this intersection,the wave packet hosts a tilted OAM that provides an additional degree of freedom to the applications that harness the OAM of photons.
Existing fusion methods struggle to maintain time performance while preserving edge and energy information. Therefore, this paper proposes an efficient multimodal medical image fusion technique based on gradient domai...
详细信息
The use of solar photovoltaic (PV) systems as a renewable source is gaining popularity as a means of reducing greenhouse gas emissions and meeting energy demand. However, losses associated with solar PV systems remain...
详细信息
The rampant spread of deepfake videos, a significant threat to media integrity, occurs widely on social media and news platforms. Detecting Deepfakes is a formidable challenge. The proposed study involves a novel algo...
详细信息
The computer vision community is embracing two promising learning paradigms: the Vision Transformer (ViT) and Multi-task Learning (MTL). ViT models show extraordinary performance over traditional convolution networks ...
The computer vision community is embracing two promising learning paradigms: the Vision Transformer (ViT) and Multi-task Learning (MTL). ViT models show extraordinary performance over traditional convolution networks but are commonly recognized as computation-intensive, especially the self-attention with quadratic complexity. MTL uses one model to infer multiple tasks with better performance by enforcing shared representation among tasks, but a huge drawback is that, most MTL regimes require activation of the entire model even when only one or a few tasks are needed, causing significant computing waste. M 3 ViT is the latest multi-task Vi $T$ model that introduces mixture-of-experts (MoE), where only a small portion of subnetworks (“experts”) are sparsely and dynamically activated based on the current task. M 3 Vi $T$ achieves better accuracy and over 80% computation reduction and paves the way for efficient real-time MTL using ViT. Despite the algorithmic advantages of MTL, ViT, and even M 3 ViT, there are still many challenges for efficient deployment on FPGA. For instance, in general Transformer/ViT models, the self-attention is known as computational intensive and requires high bandwidth. In addition, softmax operations and the activation function GELU are extensively used, which unfortunately can consume more than half of the entire FPGA resource (LUTs). In the M 3 ViT model, the promising MoE mechanism for multi-task exposes new challenges for memory access overhead and also increases resource usage because of more layer types. To address these challenges in both general Transformer/ViT models and the state-of-the-art multi-task M 3 ViT with MoE, we propose Edge-MoE, the first end-to-end FPGA accelerator for multi-task ViT with a rich collection of architectural innovations. First, for general Transformer/ViT models, we propose (1) a novel reordering mechanism for self-attention, which reduces the bandwidth requirement from proportional to constant regardl
This research addresses the basic concern of inconsistency location in control frameworks through an in-depth investigation of time arrangement examination procedures, with a particular centre on the application of Lo...
详细信息
The Energy Internet is a fundamental infrastructure for deploying green city applications, where energy saving and job acceleration are two critical issues to address. In contrast to existing approaches that focus on ...
详细信息
The Energy Internet is a fundamental infrastructure for deploying green city applications, where energy saving and job acceleration are two critical issues to address. In contrast to existing approaches that focus on static metrics with the assumption of complete prior knowledge of resource information, both application-level properties and energy-level requirements are realized in this paper by jointly considering energy saving and job acceleration during job runtime. Considering the online environment of smart city applications, the main objective is transferred as an optimization problem with a model partition and function assignment. To minimize the energy cost and job completion time together, a green workload placement approach is proposed by using the multi-action deep reinforcement learning method. Evaluations with real-world applications demonstrate the superiority of this method over state-of-the-art methods.
This study proposes a novel approach to fingerprint classification using Deep Convolutional Neural networks (CNNs), leveraging datasets from prominent sources like the National Institute of Standards and Technology (N...
详细信息
In this article, we propose an analytical method based on the adiabatic Floquet-wave expansion and the aperture field estimation technique for wide-band reflector-based leaky-wave holograms with scanning capability, r...
详细信息
暂无评论