Neural Architecture Search for Image Dehazing

• Published in: IEEE transactions on artificial intelligence Vol. 4; no. 5; pp. 1 - 11
• Main Authors: Mandal, Murari, Meedimale, Yashwanth Reddy, Reddy, M. Satish Kumar, Vipparthi, Santosh Kumar
• Format: Journal Article
• Language: English
• Published: IEEE 01.10.2023
• Subjects: Atmospheric modeling; Computer architecture; Convolutional neural networks; Feature extraction; Microprocessors; Search problems; Task analysis
• ISSN: 2691-4581; 2691-4581
• DOI: 10.1109/TAI.2022.3204732

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 network-level 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 <inline-formula><tex-math notation="LaTeX">AutoDehaze_{B}</tex-math></inline-formula>, <inline-formula><tex-math notation="LaTeX">AutoDehaze_{U1}</tex-math></inline-formula>, <inline-formula><tex-math notation="LaTeX">AutoDehaze_{U2}</tex-math></inline-formula>, and <inline-formula><tex-math notation="LaTeX">AutoDehaze_{L}</tex-math></inline-formula> which are inspired by the boat-shaped, U-shaped, 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-<inline-formula><tex-math notation="LaTeX">\beta</tex-math></inline-formula> (SOTS) and Reside-<inline-formula><tex-math notation="LaTeX">\beta</tex-math></inline-formula> (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.