Face Hallucination Revisited: An Exploratory Study on Dataset Bias

Grm, Klemen; Pernuš, Martin; Cluzel, Leo; Scheirer, Walter J.; Dobrišek, Simon

doi:10.1109/cvprw.2019.00295

Cited by 9 publications

(6 citation statements)

References 49 publications

(83 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The quality of the real-world LR is affected by a wide range of factors such as the weather, leading to the unknown complicated degradation of real LR. The gap between real LR and artificial LR is wide and results in the decrease of face super-resolution performance in [47], and needs to be bridged. It is necessary to bridge this gap.…”

Section: Gan-based Methodsmentioning

confidence: 99%

Deep Learning-based Face Super-Resolution: A Survey

Jiang¹,

Wang²,

Liu³

et al. 2021

Preprint

View full text Add to dashboard Cite

Face super-resolution, also known as face hallucination, which is aimed at enhancing the resolution of low-resolution (LR) one or a sequence of face images to generate the corresponding high-resolution (HR) face images, is a domain-specific image super-resolution problem. Recently, face super-resolution has received considerable attention, and witnessed dazzling advances with deep learning techniques. To date, few summaries of the studies on the deep learning-based face super-resolution are available. In this survey, we present a comprehensive review of deep learning techniques in face super-resolution in a systematic manner. First, we summarize the problem formulation of face super-resolution. Second, we compare the differences between generic image super-resolution and face super-resolution. Third, datasets and performance metrics commonly used in facial hallucination are presented. Fourth, we roughly categorize existing methods according to the utilization of face-specific information. In

show abstract

Section: Gan-based Methodsmentioning

confidence: 99%

Deep Learning-based Face Super-Resolution: A Survey

Jiang¹,

Wang²,

Liu³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…However, the problem with this approach is that the generated faces are often far from the true identity of the actual person, as illustrated in Figure 2. Additionally, none of the above mentioned methods are robust against noise or other corruptions in the input images [30].…”

Section: Face Super-resolutionmentioning

confidence: 99%

“…There are very few publications available in the literature which address the problem of RWSR of face‐images [30]. Furthermore, the few existing face RWSR methods are only compatible with LR images that have been squared to

16 \times 16

pixels, meaning that the reconstructed image will be only

64 \times 64

128 \times 128

pixels depending on the scaling factor [31–33].…”

Section: Related Workmentioning

confidence: 99%

Real‐world super‐resolution of face‐images from surveillance cameras

Aakerberg

Nasrollahi

Moeslund

2021

IET Image Processing

View full text Add to dashboard Cite

Most existing face image Super-Resolution (SR) methods assume that the Low-Resolution (LR) images were artificially downsampled from High-Resolution (HR) images with bicubic interpolation. This operation changes the natural image characteristics and reduces noise. Hence, SR methods trained on such data most often fail to produce good results when applied to real LR images. To solve this problem, a novel framework for the generation of realistic LR/HR training pairs is proposed. The framework estimates realistic blur kernels, noise distributions, and JPEG compression artifacts to generate LR images with similar image characteristics as the ones in the source domain. This allows to train an SR model using high-quality face images as Ground-Truth (GT). For better perceptual quality, a Generative Adversarial Network (GAN) based SR model is used, where the commonly used VGG-loss [1] is exchanged with LPIPS-loss [2]. Experimental results on both real and artificially corrupted face images show that our method results in more detailed reconstructions with less noise compared to the existing State-of-the-Art (SoTA) methods. In addition, it is shown that the traditional non-reference Image Quality Assessment (IQA) methods fail to capture this improvement and demonstrate that the more recent NIMA metric [3] correlates better with human perception via Mean Opinion Rank (MOR).

show abstract

Section: Related Workmentioning

confidence: 99%

Real-World Super-Resolution of Face-Images from Surveillance Cameras

Aakerberg¹,

Nasrollahi²,

Moeslund³

2021

Preprint

View full text Add to dashboard Cite

Most existing face image Super-Resolution (SR) methods assume that the Low-Resolution (LR) images were artificially downsampled from High-Resolution (HR) images with bicubic interpolation. This operation changes the natural image characteristics and reduces noise. Hence, SR methods trained on such data most often fail to produce good results when applied to real LR images. To solve this problem, we propose a novel framework for generation of realistic LR/HR training pairs. Our framework estimates realistic blur kernels, noise distributions, and JPEG compression artifacts to generate LR images with similar image characteristics as the ones in the source domain. This allows us to train a SR model using high quality face images as Ground-Truth (GT). For better perceptual quality we use a Generative Adversarial Network (GAN) based SR model where we have exchanged the commonly used VGG-loss [24] with LPIPS-loss [52]. Experimental results on both real and artificially corrupted face images show that our method results in more detailed reconstructions with less noise compared to existing State-of-the-Art (SoTA) methods. In addition, we show that the traditional non-reference Image Quality Assessment (IQA) methods fail to capture this improvement and demonstrate that the more recent NIMA metric [16] correlates better with human perception via Mean Opinion Rank (MOR).

show abstract

Face Hallucination Revisited: An Exploratory Study on Dataset Bias

Cited by 9 publications

References 49 publications

Deep Learning-based Face Super-Resolution: A Survey

Deep Learning-based Face Super-Resolution: A Survey

Real‐world super‐resolution of face‐images from surveillance cameras

Real-World Super-Resolution of Face-Images from Surveillance Cameras

Contact Info

Product

Resources

About