Efficient Distortion-Minimized Layerwise Pruning

Kaixin Xu

doi:10.1109/TPAMI.2025.3586418

Efficient Distortion-Minimized Layerwise Pruning

0

PRE

AI

DOI：10.1109/TPAMI.2025.3586418

Original

Original request for help

Save

Abstract

En 中文

In this paper, we propose a post-training pruning framework that jointly optimizes layerwise pruning to minimize model output distortion. Through theoretical and empirical analysis, we discover an important additivity property of output distortion from pruning weights/channels in DNNs. Leveraging this property, we reformulate pruning optimization as a combinatorial problem and solve it with dynamic programming, achieving linear time complexity and making the algorithm very fast on CPUs. Furthermore, we optimize additivity-derived distortions using Hessian-based Taylor approximation to enhance pruning efficiency, accompanied by fine-grained complexity reduction techniques. Our method is evaluated on various DNN architectures, including CNNs, ViTs, and object detectors, and on vision tasks such as image classification on CIFAR-10 and ImageNet, and 3D object detection and various datasets. We achieve SoTA with significant FLOPs reductions without accuracy loss. Specifically, on CIFAR-10, we achieve up to $27.9\times$, $29.2\times$, and $14.9\times$ FLOPs reductions on ResNet-32, VGG-16, and DenseNet-121, respectively. On ImageNet, we observe no accuracy loss with $1.69\times$ and $2\times$ FLOPs reductions on ResNet-50 and DeiT-Base, respectively. For 3D object detection, we achieve $\mathbf {3.89}\times, \mathbf {3.72}\times$ FLOPs reductions on CenterPoint and PVRCNN models. These results demonstrate the effectiveness and practicality of our approach for improving model performance through layer-adaptive weight pruning.

Keywords:

Deep learning

model compression

post-training pruning

rate-distortion theory

dynamic programming

Efficient Distortion-Minimized Layerwise Pruning

Abstract

Journal

Organization

Cited Papers