Dynamic neural network is an emerging research topic in deep learning. Compared to static
models which have fixed computational graphs and parameters at the inference stage …

Deep neural networks are generally designed as a stack of differentiable layers, in which a
prediction is obtained only after running the full stack. Recently, some contributions have …

While deeper convolutional networks are needed to achieve maximum accuracy in visual
perception tasks, for many inputs shallower networks are sufficient. We exploit this …

Artificial intelligence (AI) technologies have emerged as pivotal enablers across a multitude
of industries, including consumer electronics, healthcare, and manufacturing, largely due to …

Visual recognition is currently one of the most important and active research areas in
computer vision, pattern recognition, and even the general field of artificial intelligence. It …

Deep neural networks and decision trees operate on largely separate paradigms; typically,
the former performs representation learning with pre-specified architectures, while the latter …

Committee-based models (ensembles or cascades) construct models by combining existing
pre-trained ones. While ensembles and cascades are well-known techniques that were …

This work considers the trade-off between accuracy and testtime computational cost of deep
neural networks (DNNs) via anytime predictions from auxiliary predictions. Specifically, we …

Deep neural networks (DNNs) typically have a single exit point that makes predictions by
running the entire stack of neural layers. Since not all inputs require the same amount of …

Recently there has been a lot of progress in reducing the computation of deep models at
inference time. These methods can reduce both the computational needs and power usage …