Deploying deep learning services for time-sensitive and resource-constrained settings such as IoT using edge computing systems is a challenging task that requires dynamic adjustment of inference time. Multi-exit architectures allow deep neural networks to terminate their execution early in order to adhere to tight deadlines at the cost of accuracy. To mitigate this cost, in this paper we introduce a novel method called Multi-Exit Curriculum Learning that utilizes curriculum learning, a training strategy for neural networks that imitates human learning by sorting the training samples based on their difficulty and gradually introducing them to the network. Experiments on CIFAR-10 and CIFAR-100 datasets and various configurations of multi-exit architectures show that our method consistently improves the accuracy of early exits compared to the standard training approach.
翻译:为时间敏感和资源受限制的环境,如IoT使用边缘计算系统,提供深层次的学习服务是一项具有挑战性的任务,需要对推论时间进行动态调整,多输出结构允许深神经网络提前终止执行,以便以准确性为代价遵守紧凑的最后期限。为了降低这一成本,我们在本文件中引入了一种新的方法,即多输出课程学习,即神经网络培训战略,通过根据难度对培训样本进行分类并逐步将其引入网络,模仿人类学习。 对CIFAR-10和CIFAR-100数据集和多输出结构的各种配置进行的实验表明,与标准培训方法相比,我们的方法始终在提高早期输出的准确性。