通过 I/O 感知重构实现快速且内存高效的小波卷积

View PDF HTML (experimental)

Abstract:Wavelet convolution (WTConv) has emerged as an increasingly popular drop-in replacement for standard convolutions, expanding a network's receptive field exponentially with the number of decomposition levels while keeping the parameter count linear. However, its reference implementation is severely memory-bound due to excessive data movement through high-bandwidth memory (HBM). We develop an I/O model of WTConv to characterize this bottleneck and use it to guide three algebraic reformulations: (1) recomputing the inexpensive Haar analysis butterfly on chip, (2) collapsing the multi-level synthesis cascade into a single closed-form pass indexed by output-coordinate bits, and (3) folding learned per-channel scales into the convolution weights. Together, these reformulations enable an I/O-aware fused implementation that substantially reduces HBM traffic. We evaluate the WTConvNeXt configuration across decomposition levels and a broad range of tensor shapes. Despite performing comparable arithmetic, the reference WTConv is substantially slower than the depthwise convolution it replaces. Our reformulation reduces modeled HBM traffic by approximately $2.55\times$, yielding up to a $4.35\times$ training speedup over the reference while roughly halving peak memory usage. Thus, our reformulation preserves the benefits of WTConv while substantially reducing its execution time and memory footprint, removing the systems overhead that previously limited its practical efficiency.

Comments:	Accepted to TMLR
Subjects:	Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
Cite as:	arXiv:2608.10805 [cs.CV]
	(or arXiv:2608.10805v3 [cs.CV] for this version)
	https://doi.org/10.48550/arXiv.2608.10805 arXiv-issued DOI via DataCite

Submission history

From: Amit Aflalo [view email]
[v1] Tue, 11 Aug 2026 11:24:05 UTC (37 KB)
[v2] Thu, 13 Aug 2026 03:25:29 UTC (37 KB)
[v3] Wed, 4 Nov 2026 08:41:51 UTC (37 KB)