Scholarlink

<p>Deploying large convolutional neural networks (CNNs) on resource-constrained devices is challenging due to their high computational cost. While dynamic execution methods are promising, existing approaches for CNNs typically require specialized training or fine-tuning, limiting their effectiveness when applied to pre-trained models and requiring data access. To address this gap, we propose HASTE (<b>Has</b>hing for <b>T</b>ractable <b>E</b>fficiency), a plug-and-play convolution module that enables training-free, dynamic compression of large pre-trained CNNs. At inference time, HASTE uses locality-sensitive hashing to identify and merge redundant channels of latent feature maps on a patch-wise basis. This process simultaneously compresses the depth of both input features and their corresponding filters, resulting in computationally cheaper convolutions. We conduct extensive experiments on CIFAR-10 and ImageNet across a range of architectures, demonstrating a 46.2% FLOPs reduction in a ResNet34 on CIFAR-10 with only a 1.25% drop in accuracy, without any retraining. While current deep learning frameworks are not optimized for the dynamic, conditional operations required by our method, we provide latency estimates showing potential speedups of 1.5 <InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(\times \)</EquationSource> </InlineEquation> on CPU assuming efficient implementation of patch-wise operations. We support our claims by comprehensive ablation studies to validate our core design choices, an analysis of the method’s properties and limitations, and a discussion that connects our channel merging scheme to the conceptually related task of token merging in Vision Transformers. Our results demonstrate that HASTE provides an effective solution for steerable compression of pre-trained CNNs at runtime, opening new possibilities for the deployment of efficient deep learning methods.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

HASTE: A Framework for Training-Free, Dynamic, and Steerable Compression of Pre-trained Convolutional Neural Networks

  • 期刊论文

  • 发表日期：

    2026年6月25日

查看全文
  • Lukas Meiner,
  • Jens Mehnert,
  • Alexandru Paul Condurache

摘要

Deploying large convolutional neural networks (CNNs) on resource-constrained devices is challenging due to their high computational cost. While dynamic execution methods are promising, existing approaches for CNNs typically require specialized training or fine-tuning, limiting their effectiveness when applied to pre-trained models and requiring data access. To address this gap, we propose HASTE (Hashing for Tractable Efficiency), a plug-and-play convolution module that enables training-free, dynamic compression of large pre-trained CNNs. At inference time, HASTE uses locality-sensitive hashing to identify and merge redundant channels of latent feature maps on a patch-wise basis. This process simultaneously compresses the depth of both input features and their corresponding filters, resulting in computationally cheaper convolutions. We conduct extensive experiments on CIFAR-10 and ImageNet across a range of architectures, demonstrating a 46.2% FLOPs reduction in a ResNet34 on CIFAR-10 with only a 1.25% drop in accuracy, without any retraining. While current deep learning frameworks are not optimized for the dynamic, conditional operations required by our method, we provide latency estimates showing potential speedups of 1.5 \(\times \) on CPU assuming efficient implementation of patch-wise operations. We support our claims by comprehensive ablation studies to validate our core design choices, an analysis of the method’s properties and limitations, and a discussion that connects our channel merging scheme to the conceptually related task of token merging in Vision Transformers. Our results demonstrate that HASTE provides an effective solution for steerable compression of pre-trained CNNs at runtime, opening new possibilities for the deployment of efficient deep learning methods.