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.

通过在计算通用矩阵乘法（GEMM）时考虑激活的最终位置，我们设计了一种稀疏训练过程，以诱导可利用的半结构化激活稀疏性，并在图像分类和目标检测任务中对其进行了广泛评估，结果显示在 ImageNet 数据集上，我们的方法在 ResNet18 模型上实现了 1.25 倍的加速，并仅有 1.1% 的最小精度降低；另外，与先进的结构化剪枝方法相结合，得到的模型在延迟和准确性之间取得了很好的平衡，优于仅采用结构化剪枝技术的模型。

通过半结构化激活稀疏化加速深度神经网络