# Exploiting the Potential of Standard Convolutional Autoencoders for Image Restoration by Evolutionary Search

arXiv: https://arxiv.org/abs/1803.00370 · HTML: https://mlsanity.com/paper/1803.00370/

## Key facts

- Cited by 19 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 1 Mar 2018.

## Built on by (most cited first, 19 of 19)

- [Neural Architecture Search: Insights from 1000 Papers](https://mlsanity.com/paper/2301.08727/) (2023)
- [Efficient Architecture Search for Diverse Tasks](https://mlsanity.com/paper/2204.07554/) (2022)
- [Multi-Scale Adaptive Network for Single Image Denoising](https://mlsanity.com/paper/2203.04313/) (2022)
- [SplitNets: Designing Neural Architectures for Efficient Distributed Computing on Head-Mounted Systems](https://mlsanity.com/paper/2204.04705/) (2022)
- [GI-NAS: Boosting Gradient Inversion Attacks Through Adaptive Neural Architecture Search](https://mlsanity.com/paper/2405.20725/) (2024)
- [ISNAS-DIP: Image-Specific Neural Architecture Search for Deep Image Prior](https://mlsanity.com/paper/2111.15362/) (2021)
- [Arch-Graph: Acyclic Architecture Relation Predictor for Task-Transferable Neural Architecture Search](https://mlsanity.com/paper/2204.05941/) (2022)
- [Multi-scale Attentive Image De-raining Networks via Neural Architecture Search](https://mlsanity.com/paper/2207.00728/) (2022)
- [GP-NAS-ensemble: a model for NAS Performance Prediction](https://mlsanity.com/paper/2301.09231/) (2023)
- [Pseudo-LiDAR Based Road Detection](https://mlsanity.com/paper/2107.13279/) (2021)
- [Patching Weak Convolutional Neural Network Models through Modularization and Composition](https://mlsanity.com/paper/2209.06116/) (2022)
- [3D-ANAS: 3D Asymmetric Neural Architecture Search for Fast Hyperspectral Image Classification](https://mlsanity.com/paper/2101.04287/) (2021)
- [Data-driven Cloud Clustering via a Rotationally Invariant Autoencoder](https://mlsanity.com/paper/2103.04885/) (2021)
- [Searching Efficient Model-guided Deep Network for Image Denoising](https://mlsanity.com/paper/2104.02525/) (2021)
- [One Patch is All You Need: Joint Surface Material Reconstruction and Classification from Minimal Visual Cues](https://mlsanity.com/paper/2511.20784/) (2025)
- [Towards Regression-Free Neural Networks for Diverse Compute Platforms](https://mlsanity.com/paper/2209.13740/) (2022)
- [Single Cell Training on Architecture Search for Image Denoising](https://mlsanity.com/paper/2212.06368/) (2022)
- [Disentanglement and Compositionality of Letter Identity and Letter Position in Variational Auto-Encoder Vision Models](https://mlsanity.com/paper/2412.10446/) (2024)
- [GCA: Global Centroid Alignment in Federated Learning](https://mlsanity.com/paper/2608.22593/) (2026)

## Abstract (arXiv)

Researchers have applied deep neural networks to image restoration tasks, in which they proposed various network architectures, loss functions, and training methods. In particular, adversarial training, which is employed in recent studies, seems to be a key ingredient to success. In this paper, we show that simple convolutional autoencoders (CAEs) built upon only standard network components, i.e., convolutional layers and skip connections, can outperform the state-of-the-art methods which employ adversarial training and sophisticated loss functions. The secret is to employ an evolutionary algorithm to automatically search for good architectures. Training optimized CAEs by minimizing the ℓ2 loss between reconstructed images and their ground truths using the ADAM optimizer is all we need. Our experimental results show that this approach achieves 27.8 dB peak signal to noise ratio (PSNR) on the CelebA dataset and 40.4 dB on the SVHN dataset, compared to 22.8 dB and 33.0 dB provided by the former state-of-the-art methods, respectively.