Deep neural networks are widely used in computer vision, pattern recognition, and speech recognition and achieve high accuracy at the cost of remarkable computation. High computational complexity and memory accesses of such networks create a big challenge for using them in resource-limited and low-power embedded systems. Several binary neural networks have been proposed that exploit only 1-bit values for both weights and activations. Binary neural networks substitute complex multiply-accumulation operations with bitwise logic operations to reduce computations and memory usage. However, these quantized neural networks suffer from accuracy loss, especially in big datasets. In this paper, we introduce a quantized neural network with 2-bit weights and activations that is more accurate compared to the state-of-the-art quantized neural networks, and also the accuracy is close to the full precision neural networks. Moreover, we propose E2BNet, an efficient MAC-free hardware architecture that increases power efficiency and throughput/W about 3.6 × and 1.5 × , respectively, compared to the state-of-the-art quantized neural networks. E2BNet processes more than 500 images/s on the ImageNet dataset that not only meet real-time requirements of images/video processing but also can be deployed on high frame rate video applications.

深度神经网络广泛应用于计算机视觉、模式识别和语音识别，并以卓越的计算为代价实现高精度。这种网络的高计算复杂性和内存访问为在资源有限和低功耗嵌入式系统中使用它们带来了巨大挑战。已经提出了几个二元神经网络，它们仅利用权重和激活的 1 位值。二元神经网络用按位逻辑运算代替复杂的乘法累加运算，以减少计算和内存使用。然而，这些量化的神经网络存在精度损失，尤其是在大数据集中。在本文中，我们介绍了一个具有 2 位权重和激活的量化神经网络，与最先进的量化神经网络相比，它更准确，并且精度接近全精度神经网络。此外，我们提出了 E2BNet，这是一种高效的无 MAC 硬件架构，与最先进的量化神经网络相比，其功率效率和吞吐量/W 分别提高了约 3.6 × 和 1.5 × 。E2BNet 在 ImageNet 数据集上处理超过 500 张图像/秒，不仅满足图像/视频处理的实时性要求，而且可以部署在高帧率视频应用程序上。