PulseAugur
中
实时 10:38:05

新框架自动压缩低功耗AI视觉模型

作者 PulseAugur 编辑部 · [2 个来源] ·

研究人员开发了AQ4SViT，一个自动化的框架，旨在压缩脉冲视觉Transformer (SViTs)，以用于资源受限的嵌入式AI系统。该新框架通过采用一种利用膜电位漂移作为性能代理的搜索门控策略，解决了手动量化的可扩展性问题。AQ4SViT提供两种搜索变体：贪婪搜索，速度更快但可能找到局部最优解；束搜索，速度较慢但旨在找到全局最优解。实验表明，AQ4SViT-Greedy可实现高达82.5%的内存节省，AQ4SViT-Beam可实现高达90%的内存节省，同时保持高精度。 AI

影响 该框架可以实现更高效的AI模型在边缘设备上的部署，扩展嵌入式AI系统的能力。

排序理由 该集群包含一篇在arXiv上发表的研究论文，详细介绍了一个新的模型压缩框架。

在 arXiv cs.NE (Neural & Evolutionary) 阅读 →

AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →

报道来源 [2]

  1. arXiv cs.AI TIER_1 English(EN) · Rachmad Vidya Wicaksana Putra, Saad Iftikhar, Muhammad Shafique ·

    AQ4SViT: An Automated Quantization Framework with Search Gating Policy for Compressing Spiking Vision Transformers

    arXiv:2606.15523v1 Announce Type: cross Abstract: Spiking Vision Transformers (SViTs) have emerged as alternative low-power ViT models, but their large sizes hinder their deployments on resource-constrained embedded AI systems. To address this, state-of-the-art works proposed qua…

  2. arXiv cs.NE (Neural & Evolutionary) TIER_1 English(EN) · Muhammad Shafique ·

    AQ4SViT: An Automated Quantization Framework with Search Gating Policy for Compressing Spiking Vision Transformers

    Spiking Vision Transformers (SViTs) have emerged as alternative low-power ViT models, but their large sizes hinder their deployments on resource-constrained embedded AI systems. To address this, state-of-the-art works proposed quantization techniques to compress SViT models, but …

相关实体

相关话题