# Accelerating Deep Neural Networks via Semi-Structured Activation Sparsity

Source: [Accelerating Deep Neural Networks via Semi-Structured Activation Sparsity](<https://www.sota2.com/research/paper/68fa70ae1e209131cea4c7cd>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-26T01:46:58.939Z

Authors: Matteo Grimaldi, Darshan C. Ganji, Ivan Lazarevich, Sudhakar Sah

Published: 2023-09-12T22:28:53.000Z

## Abstract

The demand for efficient processing of deep neural networks (DNNs) on embedded devices is a significant challenge limiting their deployment. Exploiting sparsity in the network's feature maps is one of the ways to reduce its inference latency. It is known that unstructured sparsity results in lower accuracy degradation with respect to structured sparsity but the former needs extensive inference engine changes to get latency benefits. To tackle this challenge, we propose a solution to induce semi-structured activation sparsity exploitable through minor runtime modifications. To attain high speedup levels at inference time, we design a sparse training procedure with awareness of the final position of the activations while computing the General Matrix Multiplication (GEMM). We extensively evaluate the proposed solution across various models for image classification and object detection tasks. Remarkably, our approach yields a speed improvement of $1.25 \\times$ with a minimal accuracy drop of $1.1\\%$ for the ResNet18 model on the ImageNet dataset. Furthermore, when combined with a state-of-the-art structured pruning method, the resulting models provide a good latency-accuracy trade-off, outperforming models that solely employ structured pruning techniques.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2309.06626>)

[PDF](<https://arxiv.org/pdf/2309.06626.pdf>)
