E$^2$CM: Early Exit via Class Means for Efficient Supervised and Unsupervised Learning
This work addresses efficiency challenges in neural network training for low-power devices, such as in wireless edge networks, offering an incremental improvement over existing early exit methods.
The paper tackles the problem of high training costs in early exit mechanisms for neural networks by proposing E^2CM, a technique based on class means that avoids gradient-based training and network modifications. The results show that E^2CM achieves higher accuracy than existing methods under fixed training time budgets and can boost performance when combined with other schemes, with evaluations on datasets like CIFAR-100 and ImageNet.
State-of-the-art neural networks with early exit mechanisms often need considerable amount of training and fine tuning to achieve good performance with low computational cost. We propose a novel early exit technique, Early Exit Class Means (E$^2$CM), based on class means of samples. Unlike most existing schemes, E$^2$CM does not require gradient-based training of internal classifiers and it does not modify the base network by any means. This makes it particularly useful for neural network training in low-power devices, as in wireless edge networks. We evaluate the performance and overheads of E$^2$CM over various base neural networks such as MobileNetV3, EfficientNet, ResNet, and datasets such as CIFAR-100, ImageNet, and KMNIST. Our results show that, given a fixed training time budget, E$^2$CM achieves higher accuracy as compared to existing early exit mechanisms. Moreover, if there are no limitations on the training time budget, E$^2$CM can be combined with an existing early exit scheme to boost the latter's performance, achieving a better trade-off between computational cost and network accuracy. We also show that E$^2$CM can be used to decrease the computational cost in unsupervised learning tasks.