Osman Furkan Kınlı

Auto-white balance (AWB) correction is a critical operation in image signal processors (ISPs) for accurate and consistent color correction across various illumination scenarios. This paper presents a novel and efficient AWB correction method that achieves at least 35 times faster processing with equivalent or superior performance on high-resolution images for the current state-of-the-art methods. Inspired by deterministic color style transfer, our approach introduces deterministic illumination color mapping, leveraging learnable projection matrices for both canonical illumination form and AWB-corrected output. It involves feeding high-resolution images and corresponding latent representations into a mapping module to derive a canonical form, followed by another mapping module that maps the pixel values to those for the corrected version. This strategy is designed as resolution-agnostic and enables seamless integration of any pre-trained AWB network as the backbone. Our method provides an efficient deep learning-based AWB correction solution, promising real-time, high-quality color correction for digital imaging applications.

In this reproducibility study, we present our results and experience while replicating the paper titled Exact Feature Distribution Matching for Arbitrary Style Transfer and Domain Generalization. In real‐world scenarios, the feature distributions are mostly much more complicated than Gaussian, so only mean and standard deviation may not be fully representative to match them. This paper introduces a novel strategy to exactly match the histograms of image features via the sort‐matching algorithm in a computationally feasible way. We were able to reproduce most of the results presented in the original paper both qualitatively and quantitatively.

Style may refer to different concepts (e.g., painting style, hairstyle, texture, color, filter, etc.) depending on how the feature space is formed. In this work, we propose a novel idea of interpreting the lighting in the single- and multi-illuminant scenes as the concept of style. To verify this idea, we introduce an enhanced auto white-balance (AWB) method that models the lighting in single- and mixed-illuminant scenes as the style factor. Our AWB method does not require any illumination estimation step, yet it contains a network learning to generate the weighting maps of the images with different WB settings. The proposed network utilizes the style information extracted from the scene by a multi-head style extraction module. AWB correction is completed after blending these weighting maps and the scene. Experiments on single- and mixed illuminant datasets demonstrate that our proposed method achieves promising correction results when compared to the recent works. This shows that the lighting in the scenes with multiple illuminations can be modeled by the
concept of style.

Image-level corruptions and perturbations degrade the performance of CNNs on different downstream vision tasks. Social media filters are one of the most common sources of corruptions and perturbations for real-world visual analysis applications. The negative effects of these distractive factors can be alleviated by recovering the original images with their pure style for the inference of the downstream vision tasks. Assuming these filters substantially inject a piece of additional style information into the social media images, we can formulate the problem of recovering the original versions as a reverse style transfer problem. We introduce the Contrastive Instagram Filter Removal Network (CIFR), which enhances this idea for Instagram filter removal by employing a novel multi-layer patch-wise contrastive style learning mechanism. Experiments show our proposed strategy produces better qualitative and quantitative results than the previous studies. Finally, we present the inference outputs and quantitative comparison of filtered and recovered images on localization and segmentation tasks to encourage the main motivation for this problem.

Social media images are generally transformed by filtering to obtain an aesthetically more pleasing appearance, however, convolutional neural networks (CNNs) generally fail to interpret both the image and its filtered version as the same in the visual analysis of social media images.

We introduced the Instagram Filter Removal Network (IFRNet) to mitigate the effects of image filters for social media analysis applications. To achieve this, we assumed any filter applied to an image substantially injects a piece of additional style information to it, and we considered this problem as a reverse style transfer problem.

The visual effects of filtering can be directly removed by adaptively normalizing external style information in each encoder level. Experiments demonstrate that IFRNet outperforms all compared methods in quantitative and qualitative comparisons and can remove the visual effects to a great extent. Additionally, we present the filter classification performance of our proposed model and analyze the dominant color estimation on the images unfiltered by all compared methods.

Fashion image understanding is an active research field with a large number of practical applications for the industry and despite its practical impacts on intelligent fashion analysis systems, clothing image inpainting has not been extensively examined yet; for that matter, we presented an extensive benchmark of clothing image inpainting on well-known fashion datasets.

Furthermore, we introduced the use of a dilated version of partial convolutions, which efficiently derive the mask update step, and empirically show that the proposed method reduces the required number of layers to form fully-transparent masks. Experiments show that dilated partial convolutions (DPConv) improve the quantitative inpainting performance compared to the other inpainting strategies; it performs better when the mask size is 20% or more of the image.

You can find more information in the following paper: https://arxiv.org/pdf/2007.05080.pdf.

Inpainting a particular missing region in an image is a challenging vision task, and promising improvements on this task have been achieved with the help of the recent developments in vision-related deep learning studies. Although it may directly impact the decisions of AI-based fashion analysis systems, a limited number of studies for image inpainting have been done in the fashion domain.

This study proposes a multi-modal, generative deep learning approach for filling the missing parts in fashion images by constraining visual features with textual features extracted from image descriptions. Our model comprises four main blocks, which can be introduced as textual feature extractor, coarse image generator guided by textual features, fine image generator enhancing the coarse output, and global and local discriminators improving refined outputs.

Several experiments conducted on the FashionGen dataset with different combinations of neural network components show that our multi-modal approach can generate visually plausible patches to fill the missing parts in the images.

In this study, we investigate the in-shop clothing retrieval performance of densely-connected capsule networks with dynamic routing. To achieve this, we propose a triplet-based design of capsule network architecture with two different feature extraction methods. In our design, stacked convolutional (SC) and residual-connected (RC) blocks are used to form the input of capsule layers.

Experimental results show that both of our designs outperform all variants of the baseline study, namely FashionNet, without relying on the landmark information. Moreover, compared to the SOTA architectures on clothing retrieval, our proposed triplet capsule networks achieve comparable recall rates only with half of the parameters used in the SOTA architectures.