注意力卷积：统一自注意力的表现力与卷积的效率
Attentive Convolution: Unifying the Expressivity of Self-Attention with Convolutional Efficiency
AI总结：
本文提出注意力卷积（ATConv），通过自适应路由和侧抑制原则重新设计卷积算子，仅用3×3内核即可超越多种自注意力机制，在ImageNet-1K上以27M参数达到84.4%准确率，并提升扩散模型生成性能。
摘要：
自注意力（SA）因其相对于传统卷积（Conv）的强大表现力，已成为现代视觉骨干网络的基石。然而，其二次复杂度仍然是实际应用中的关键瓶颈。鉴于卷积具有线性复杂度和强大的视觉先验，人们不断努力推动卷积的复兴。然而，性能差距依然存在，这表明这些现代化改造尚未捕捉到定义SA的内在表现力。在本文中，我们重新审视CNN的设计，围绕一个关键问题展开：是什么原则使SA优于Conv？结果，我们揭示了两个基本见解，挑战了先前研究中长期存在的设计直觉（例如，感受野）。这两个发现是：（1）自适应路由：SA根据语义内容动态调节位置信息流，而Conv在所有位置使用静态内核。（2）侧抑制：SA在令牌加权中引发分数竞争，有效抑制冗余并锐化表示，而Conv滤波器缺乏这种抑制动态，表现出相当大的冗余。基于此，我们提出了注意力卷积（ATConv），这是一种对卷积算子的原则性重新表述，内在地注入了这些原则。有趣的是，仅使用3×3内核，ATConv在基本视觉任务中持续优于各种SA机制。基于ATConv，我们引入了AttNet，一个CNN家族，仅用27M参数即可达到84.4%的ImageNet-1K Top-1准确率。在基于扩散的图像生成中，将SiT-XL/2中的所有SA替换为所提出的3×3 ATConv，在400k步内将ImageNet FID降低0.15，且采样更快。代码可在github.com/price112/Attentive-Convolution获取。
英文摘要：
Self-attention (SA) has become the cornerstone of modern vision backbones for its powerful expressivity over traditional Convolutions (Conv). However, its quadratic complexity remains a critical bottleneck for practical applications. Given that Conv offers linear complexity and strong visual priors, continuing efforts have been made to promote the renaissance of Conv. However, a persistent performance chasm remains, highlighting that these modernizations have not yet captured the intrinsic expressivity that defines SA. In this paper, we re-examine the design of the CNNs, directed by a key question: what principles give SA its edge over Conv? As a result, we reveal two fundamental insights that challenge the long-standing design intuitions in prior research (e.g., Receptive field). The two findings are: (1) \textit{Adaptive routing}: SA dynamically regulates positional information flow according to semantic content, whereas Conv employs static kernels uniformly across all positions. (2) \textit{Lateral inhibition}: SA induces score competition among token weighting, effectively suppressing redundancy and sharpening representations, whereas Conv filters lack such inhibitory dynamics and exhibit considerable redundancy. Based on this, we propose \textit{Attentive Convolution} (ATConv), a principled reformulation of the convolutional operator that intrinsically injects these principles. Interestingly, with only $3\times3$ kernels, ATConv consistently outperforms various SA mechanisms in fundamental vision tasks. Building on ATConv, we introduce AttNet, a CNN family that can attain \textbf{84.4\%} ImageNet-1K Top-1 accuracy with only 27M parameters. In diffusion-based image generation, replacing all SA with the proposed $3\times 3$ ATConv in SiT-XL/2 reduces ImageNet FID by 0.15 in 400k steps with faster sampling. Code is available at: github.com/price112/Attentive-Convolution.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
