Abstract:Identifying small size images or small objects is a notoriously challenging problem, as discriminative representations are difficult to learn from the limited information contained in them with poor-quality appearance and unclear object structure. Existing research works usually increase the resolution of low-resolution image in the pixel space in order to provide better visual quality for human viewing. However, the improved performance of such methods is usually limited or even trivial in the case of very small image size (we will show it in this paper explicitly).In this paper, different from image super-resolution (IS-R), we propose a novel super-resolution technique called feature super-resolution (FSR), which aims at enhancing the discriminatory power of small size image in order to provide high recognition precision for machine. To achieve this goal, we propose a new Feature Super-Resolution Generative Adversarial Network (FSR-GAN) model that transforms the raw poor features of small size images to highly discriminative ones by performing super-resolution in the feature space. Our FSR-GAN consists of two subnetworks: a feature generator network G and a feature discriminator network D. By training the G and the D networks in an alternative manner, we encourage the G network to discover the latent distribution correlations between small size and large size images and then use G to improve the representations of small images. Extensive experiment results on Oxford5K, Paris, Holidays, and Flick100k datasets demonstrate that the proposed FSR approach can effectively enhance the discriminatory ability of features. Even when the resolution of query images is reduced greatly, e.g., 1/64 original size, the query feature enhanced by our FSR approach achieves surprisingly high retrieval performance at different image resolutions and increases the retrieval precision by 25% compared to the raw query feature.[GRAPHICS].

Feature-based Recognition Framework for Super-resolution Images

Super-resolution Reconstruction Algorithms Based on Fusion of Deep Learning Mechanism and Wavelet

Single Remote Sensing Image Super-Resolution Via a Generative Adversarial Network with Stratified Dense Sampling and Chain Training

Residual Feature Aggregation Network for Image Super-Resolution.

Single-image Super-Resolution Via Selective Multi-Scale Network

Feature Super-Resolution: Make Machine See More Clearly

A Multi-Scale Recursive Attention Feature Fusion Network for Image Super-Resolution Reconstruction Algorithm

MFFAGAN: Generative Adversarial Network With Multilevel Feature Fusion Attention Mechanism for Remote Sensing Image Super-Resolution

FSR: A General Frequency-Oriented Framework to Accelerate Image Super-resolution Networks.

MFAAnet: New Feature Extraction Network in Image Super-Resolution

Improved Residual Dense Network for Large Scale Super-Resolution via Generative Adversarial Network

Remote Sensing Image Super-Resolution Adversarial Network Based on Reverse Feature Fusion and Residual Feature Dilation

Attention-Guided Multi-scale Interaction Network for Face Super-Resolution

A Progressive Feature Enhancement Deep Network for Large-Scale Remote Sensing Image Superresolution

IESRGAN: Enhanced U-Net Structured Generative Adversarial Network for Remote Sensing Image Super-Resolution Reconstruction

Multi-Frame Super-Resolution Algorithm Based on a WGAN

Feature Super-Resolution Based Facial Expression Recognition for Multi-Scale Low-Resolution Images

An improved generative adversarial networks for remote sensing image super-resolution reconstruction via multi-scale residual block

Lightweight hierarchical residual feature fusion network for single-image super-resolution

Forest Single-Frame Remote Sensing Image Super-Resolution Using GANs

EFRG-SRGAN: combining augmented features for real-world super-resolution