Abstract:
Manual design of deep networks require numerous
trials and parameter tuning, resulting in inefficient utilization
of time, energy, and resources. In this work, we present a
neural architecture search (NAS) algorithm - AutoDehaze, to
automatically discover effective neural network for single image
dehazing. The proposed AutoDehaze algorithm is built on the
the gradient based search strategy and hierarchical networklevel optimization. We construct a set of search space layouts
to reduce memory consumption, avoid the NAS collapse issue,
and considerably accelerate the search speed. We propose four
search spaces AutoDehazeB, AutoDehazeU1, AutoDehazeU2,
and AutoDehazeL which are inspired by the boat-shaped, Ushaped, and lateral connection-based designs. To the best of our
knowledge, this is a first attempt to present a NAS method
for dehazing with a variety of network search strategies. We
conduct a comprehensive set of experiments on Reside-Standard
(SOTS), Reside-β (SOTS) and Reside-β (HSTS), D-Hazy, and
HazeRD datasets. The architectures discovered by the proposed
AutoDehaze quantitatively and qualitatively outperform the existing state-of-the-art approaches. The experiments also show that
our models have considerably fewer parameters and runs at a
faster inference speed in both CPU and GPU devices.
