IamNN: Iterative and Adaptive Mobile Neural Network for Efficient Image Classification

04/26/2018
by   Sam Leroux, et al.
0

Deep residual networks (ResNets) made a recent breakthrough in deep learning. The core idea of ResNets is to have shortcut connections between layers that allow the network to be much deeper while still being easy to optimize avoiding vanishing gradients. These shortcut connections have interesting side-effects that make ResNets behave differently from other typical network architectures. In this work we use these properties to design a network based on a ResNet but with parameter sharing and with adaptive computation time. The resulting network is much smaller than the original network and can adapt the computational cost to the complexity of the input image.

READ FULL TEXT
research
09/18/2017

Wide and deep volumetric residual networks for volumetric image classification

3D shape models that directly classify objects from 3D information have ...
research
10/30/2021

Neural Network based on Automatic Differentiation Transformation of Numeric Iterate-to-Fixedpoint

This work proposes a Neural Network model that can control its depth usi...
research
06/01/2017

DiracNets: Training Very Deep Neural Networks Without Skip-Connections

Deep neural networks with skip-connections, such as ResNet, show excelle...
research
04/30/2018

Towards Deeper Generative Architectures for GANs using Dense connections

In this paper, we present the result of adopting skip connections and de...
research
12/07/2016

Spatially Adaptive Computation Time for Residual Networks

This paper proposes a deep learning architecture based on Residual Netwo...
research
10/06/2021

SIRe-Networks: Skip Connections over Interlaced Multi-Task Learning and Residual Connections for Structure Preserving Object Classification

Improving existing neural network architectures can involve several desi...
research
12/14/2021

An Interpretive Constrained Linear Model for ResNet and MgNet

We propose a constrained linear data-feature-mapping model as an interpr...

Please sign up or login with your details

Forgot password? Click here to reset