Accuracy Evaluation of Transposed Convolution-Based Quantized Neural Networks
Accuracy Evaluation of Transposed Convolution-Based Quantized Neural Networks
复制标题
DOI：
10.1109/ijcnn55064.2022.9892671
复制
发表时间：
2022-07
期刊：
影响因子：
--
通讯作者：
Cristian Sestito;S. Perri;Rob Stewart
中科院分区：
文献类型：
--
作者：
Cristian Sestito;S. Perri;Rob Stewart
人工智能领域的一些现代应用利用深度学习来做出准确的决策。最近关于压缩技术的工作允许深度学习应用程序（如计算机视觉）在边缘计算设备上运行。例如，量化深度学习架构的精度允许边缘计算设备以低功耗实现高吞吐量。量化主要集中在多层感知器和基于卷积的分类问题模型上。然而，它对更复杂场景（如图像上采样）的影响仍未得到充分探索。本文系统地评估了量化神经网络在图像压缩/解压缩、合成图像生成和语义分割三种不同应用中进行图像上采样时所达到的精度。考虑到可学习滤波器预测像素的前景，采用转置卷积层进行上采样。基于分析度量的实验结果表明，当量化范围在3 ~ 7位之间时，达到了可接受的精度。根据目视检查，2-6位范围保证了适当的精度。
Several modern applications in the field of Artificial Intelligence exploit deep learning to make accurate decisions. Recent work on compression techniques allows for deep learning applications, such as computer vision, to run on Edge Computing devices. For instance, quantizing the precision of deep learning architectures allows Edge Computing devices to achieve high throughput at low power. Quantization has been mainly focused on multilayer perceptrons and convolution-based models for classification problems. However, its impact over more complex scenarios, such as image up-sampling, is still underexplored. This paper presents a systematic evaluation of the accuracy achieved by quantized neural networks when performing image up-sampling in three different applications: image compression/decompression, synthetic image generation and semantic segmentation. Taking into account the promising attitude of learnable filters to predict pixels, transposed convolutional layers are used for up-sampling. Experimental results based on analytical metrics show that acceptable accuracies are reached with quantization spanning between 3 and 7 bits. Based on the visual inspection, the range 2–6 bits guarantees appropriate accuracy.