Gozde Unal

Gulcin Baykal, Melih Kandemir, Gozde Unal

Codebook collapse is a common problem in training deep generative models with discrete representation spaces like Vector Quantized Variational Autoencoders (VQ-VAEs). We observe that the same problem arises for the alternatively designed discrete variational autoencoders (dVAEs) whose encoder directly learns a distribution over the codebook embeddings to represent the data. We hypothesize that using the softmax function to obtain a probability distribution causes the codebook collapse by assigning overconfident probabilities to the best matching codebook elements. In this paper, we propose a novel way to incorporate evidential deep learning (EDL) instead of softmax to combat the codebook collapse problem of dVAE. We evidentially monitor the significance of attaining the probability distribution over the codebook embeddings, in contrast to softmax usage. Our experiments using various datasets show that our model, called EdVAE, mitigates codebook collapse while improving the reconstruction performance, and enhances the codebook usage compared to dVAE and VQ-VAE based models. Our code can be found at https://github.com/ituvisionlab/EdVAE .

Gulcin Baykal, Melih Kandemir, Gozde Unal

Disentangled representation learning aims to represent the underlying generative factors of a dataset in a latent representation independently of one another. In our work, we propose a discrete variational autoencoder (VAE) based model where the ground truth information about the generative factors are not provided to the model. We demonstrate the advantages of learning discrete representations over learning continuous representations in facilitating disentanglement. Furthermore, we propose incorporating an inductive bias into the model to further enhance disentanglement. Precisely, we propose scalar quantization of the latent variables in a latent representation with scalar values from a global codebook, and we add a total correlation term to the optimization as an inductive bias. Our method called FactorQVAE combines optimization based disentanglement approaches with discrete representation learning, and it outperforms the former disentanglement methods in terms of two disentanglement metrics (DCI and InfoMEC) while improving the reconstruction performance. Our code can be found at https://github.com/ituvisionlab/FactorQVAE.

Alper Unal, Busra Asan, Ismail Sezen et al.

Understanding seasonal climatic conditions is critical for better management of resources such as water, energy and agriculture. Recently, there has been a great interest in utilizing the power of artificial intelligence methods in climate studies. This paper presents a cutting-edge deep learning model (UNet++) trained by state-of-the-art global CMIP6 models to forecast global temperatures a month ahead using the ERA5 reanalysis dataset. ERA5 dataset was also used for finetuning as well performance analysis in the validation dataset. Three different setups (CMIP6; CMIP6 + elevation; CMIP6 + elevation + ERA5 finetuning) were used with both UNet and UNet++ algorithms resulting in six different models. For each model 14 different sequential and non-sequential temporal settings were used. The Mean Absolute Error (MAE) analysis revealed that UNet++ with CMIP6 with elevation and ERA5 finetuning model with "Year 3 Month 2" temporal case provided the best outcome with an MAE of 0.7. Regression analysis over the validation dataset between the ERA5 data values and the corresponding AI model predictions revealed slope and $R^2$ values close to 1 suggesting a very good agreement. The AI model predicts significantly better than the mean CMIP6 ensemble between 2016 and 2021. Both models predict the summer months more accurately than the winter months.

Halil Faruk Karagoz, Gulcin Baykal, Irem Arikan Eksi et al.

The problem of text-guided image generation is a complex task in Computer Vision, with various applications, including creating visually appealing artwork and realistic product images. One popular solution widely used for this task is the diffusion model, a generative model that generates images through an iterative process. Although diffusion models have demonstrated promising results for various image generation tasks, they may only sometimes produce satisfactory results when applied to more specific domains, such as the generation of textile patterns based on text guidance. This study presents a fine-tuned diffusion model specifically trained for textile pattern generation by text guidance to address this issue. The study involves the collection of various textile pattern images and their captioning with the help of another AI model. The fine-tuned diffusion model is trained with this newly created dataset, and its results are compared with the baseline models visually and numerically. The results demonstrate that the proposed fine-tuned diffusion model outperforms the baseline models in terms of pattern quality and efficiency in textile pattern generation by text guidance. This study presents a promising solution to the problem of text-guided textile pattern generation and has the potential to simplify the design process within the textile industry.

V. Bugra Yesilkaynak, Emine Dari, Alican Mertan et al.

Existing multi-label frameworks only exploit the information deduced from the bipartition of the labels into a positive and negative set. Therefore, they do not benefit from the ranking order between positive labels, which is the concept we introduce in this paper. We propose a novel multi-label ranking method: GaussianMLR, which aims to learn implicit class significance values that determine the positive label ranks instead of treating them as of equal importance, by following an approach that unifies ranking and classification tasks associated with multi-label ranking. Due to the scarcity of public datasets, we introduce eight synthetic datasets generated under varying importance factors to provide an enriched and controllable experimental environment for this study. On both real-world and synthetic datasets, we carry out extensive comparisons with relevant baselines and evaluate the performance on both of the two sub-tasks. We show that our method is able to accurately learn a representation of the incorporated positive rank order, which is not only consistent with the ground truth but also proportional to the underlying information. We strengthen our claims empirically by conducting comprehensive experimental studies. Code is available at https://github.com/MrGranddy/GaussianMLR.

Abdullah Akgül, Gozde Unal, Melih Kandemir

We study the problem of fitting a model to a dynamical environment when new modes of behavior emerge sequentially. The learning model is aware when a new mode appears, but it cannot access the true modes of individual training sequences. The state-of-the-art continual learning approaches cannot handle this setup, because parameter transfer suffers from catastrophic interference and episodic memory design requires the knowledge of the ground-truth modes of sequences. We devise a novel continual learning method that overcomes both limitations by maintaining a \textit{descriptor} of the mode of an encountered sequence in a neural episodic memory. We employ a Dirichlet Process prior on the attention weights of the memory to foster efficient storage of the mode descriptors. Our method performs continual learning by transferring knowledge across tasks by retrieving the descriptors of similar modes of past tasks to the mode of a current sequence and feeding this descriptor into its transition kernel as control input. We observe the continual learning performance of our method to compare favorably to the mainstream parameter transfer approach.

Gulcin Baykal, Halil Faruk Karagoz, Taha Binhuraib et al.

Diffusion models are generative models that have shown significant advantages compared to other generative models in terms of higher generation quality and more stable training. However, the computational need for training diffusion models is considerably increased. In this work, we incorporate prototype learning into diffusion models to achieve high generation quality faster than the original diffusion model. Instead of randomly initialized class embeddings, we use separately learned class prototypes as the conditioning information to guide the diffusion process. We observe that our method, called ProtoDiffusion, achieves better performance in the early stages of training compared to the baseline method, signifying that using the learned prototypes shortens the training time. We demonstrate the performance of ProtoDiffusion using various datasets and experimental settings, achieving the best performance in shorter times across all settings.

Atahan Ozer, Kadir Burak Buldu, Abdullah Akgül et al.

Federated Learning enables multiple data centers to train a central model collaboratively without exposing any confidential data. Even though deterministic models are capable of performing high prediction accuracy, their lack of calibration and capability to quantify uncertainty is problematic for safety-critical applications. Different from deterministic models, probabilistic models such as Bayesian neural networks are relatively well-calibrated and able to quantify uncertainty alongside their competitive prediction accuracy. Both of the approaches appear in the federated learning framework; however, the aggregation scheme of deterministic models cannot be directly applied to probabilistic models since weights correspond to distributions instead of point estimates. In this work, we study the effects of various aggregation schemes for variational Bayesian neural networks. With empirical results on three image classification datasets, we observe that the degree of spread for an aggregated distribution is a significant factor in the learning process. Hence, we present an investigation on the question of how to combine variational Bayesian networks in federated learning, while providing benchmarks for different aggregation settings.

M. Arda Aydın, Efe Mert Çırpar, Elvin Abdinli et al.

Recent advances in foundational Vision Language Models (VLMs) have reshaped the evaluation paradigm in computer vision tasks. These foundational models, especially CLIP, have accelerated research in open-vocabulary computer vision tasks, including Open-Vocabulary Semantic Segmentation (OVSS). Although the initial results are promising, the dense prediction capabilities of VLMs still require further improvement. In this study, we enhance the semantic segmentation performance of CLIP by introducing new modules and modifications: 1) architectural changes in the last layer of ViT and the incorporation of attention maps from the middle layers with the last layer, 2) Image Engineering: applying data augmentations to enrich input image representations, and 3) using Large Language Models (LLMs) to generate definitions and synonyms for each class name to leverage CLIP's open-vocabulary capabilities. Our training-free method, ITACLIP, outperforms current state-of-the-art approaches on segmentation benchmarks such as COCO-Stuff, COCO-Object, Pascal Context, and Pascal VOC. Our code is available at https://github.com/m-arda-aydn/ITACLIP.

Sevgi Altun, Mustafa Cem Gunes, Yusuf H. Sahin et al.

This study integrates artificial intelligence and computational design tools to extract information from architectural heritage. Photogrammetry-based point cloud models of brick walls from the Anatolian Seljuk period are analysed in terms of the interrelated units of construction, simultaneously considering both the inherent symmetries and irregularities. The real-world data is used as input for acquiring the stochastic parameters of spatial relations and a set of parametric shape rules to recreate designs of existing and hypothetical brick walls within the style. The motivation is to be able to generate large data sets for machine learning of the style and to devise procedures for robotic production of such designs with repetitive units.

Mert Gulsen, Batuhan Cengiz, Yusuf H. Sahin et al.

Point clouds are extensively employed in a variety of real-world applications such as robotics, autonomous driving and augmented reality. Despite the recent success of point cloud neural networks, especially for safety-critical tasks, it is essential to also ensure the robustness of the model. A typical way to assess a model's robustness is through adversarial attacks, where test-time examples are generated based on gradients to deceive the model. While many different defense mechanisms are studied in 2D, studies on 3D point clouds have been relatively limited in the academic field. Inspired from PointDP, which denoises the network inputs by diffusion, we propose Point Cloud Layerwise Diffusion (PCLD), a layerwise diffusion based 3D point cloud defense strategy. Unlike PointDP, we propagated the diffusion denoising after each layer to incrementally enhance the results. We apply our defense method to different types of commonly used point cloud models and adversarial attacks to evaluate its robustness. Our experiments demonstrate that the proposed defense method achieved results that are comparable to or surpass those of existing methodologies, establishing robustness through a novel technique. Code is available at https://github.com/batuceng/diffusion-layer-robustness-pc.

Emine Dari, V. Bugra Yesilkaynak, Alican Mertan et al.

Multi-label ranking maps instances to a ranked set of predicted labels from multiple possible classes. The ranking approach for multi-label learning problems received attention for its success in multi-label classification, with one of the well-known approaches being pairwise label ranking. However, most existing methods assume that only partial information about the preference relation is known, which is inferred from the partition of labels into a positive and negative set, then treat labels with equal importance. In this paper, we focus on the unique challenge of ranking when the order of the true label set is provided. We propose a novel dedicated loss function to optimize models by incorporating penalties for incorrectly ranked pairs, and make use of the ranking information present in the input. Our method achieves the best reported performance measures on both synthetic and real world ranked datasets and shows improvements on overall ranking of labels. Our experimental results demonstrate that our approach is generalizable to a variety of multi-label classification and ranking tasks, while revealing a calibration towards a certain ranking ordering.

Batuhan Cengiz, Mert Gulsen, Yusuf H. Sahin et al.

Point clouds and meshes are widely used 3D data structures for many computer vision applications. While the meshes represent the surfaces of an object, point cloud represents sampled points from the surface which is also the output of modern sensors such as LiDAR and RGB-D cameras. Due to the wide application area of point clouds and the recent advancements in deep neural networks, studies focusing on robust classification of the 3D point cloud data emerged. To evaluate the robustness of deep classifier networks, a common method is to use adversarial attacks where the gradient direction is followed to change the input slightly. The previous studies on adversarial attacks are generally evaluated on point clouds of daily objects. However, considering 3D faces, these adversarial attacks tend to affect the person's facial structure more than the desired amount and cause malformation. Specifically for facial expressions, even a small adversarial attack can have a significant effect on the face structure. In this paper, we suggest an adversarial attack called $ε$-Mesh Attack, which operates on point cloud data via limiting perturbations to be on the mesh surface. We also parameterize our attack by $ε$ to scale the perturbation mesh. Our surface-based attack has tighter perturbation bounds compared to $L_2$ and $L_\infty$ norm bounded attacks that operate on unit-ball. Even though our method has additional constraints, our experiments on CoMA, Bosphorus and FaceWarehouse datasets show that $ε$-Mesh Attack (Perpendicular) successfully confuses trained DGCNN and PointNet models $99.72\%$ and $97.06\%$ of the time, with indistinguishable facial deformations. The code is available at https://github.com/batuceng/e-mesh-attack.

Gulcin Baykal, Furkan Ozcelik, Gozde Unal

Generative Adversarial Networks (GANs) have become the most used networks towards solving the problem of image generation. Self-supervised GANs are later proposed to avoid the catastrophic forgetting of the discriminator and to improve the image generation quality without needing the class labels. However, the generalizability of the self-supervision tasks on different GAN architectures is not studied before. To that end, we extensively analyze the contribution of a previously proposed self-supervision task, deshuffling of the DeshuffleGANs in the generalizability context. We assign the deshuffling task to two different GAN discriminators and study the effects of the task on both architectures. We extend the evaluations compared to the previously proposed DeshuffleGANs on various datasets. We show that the DeshuffleGAN obtains the best FID results for several datasets compared to the other self-supervised GANs. Furthermore, we compare the deshuffling with the rotation prediction that is firstly deployed to the GAN training and demonstrate that its contribution exceeds the rotation prediction. We design the conditional DeshuffleGAN called cDeshuffleGAN to evaluate the quality of the learnt representations. Lastly, we show the contribution of the self-supervision tasks to the GAN training on the loss landscape and present that the effects of these tasks may not be cooperative to the adversarial training in some settings. Our code can be found at https://github.com/gulcinbaykal/DeshuffleGAN.

V. Bugra Yesilkaynak, Emine Dari, Alican Mertan et al.

Existing multi-label ranking (MLR) frameworks only exploit information deduced from the bipartition of labels into positive and negative sets. Therefore, they do not benefit from ranking among positive labels, which is the novel MLR approach we introduce in this paper. We propose UniMLR, a new MLR paradigm that models implicit class relevance/significance values as probability distributions using the ranking among positive labels, rather than treating them as equally important. This approach unifies ranking and classification tasks associated with MLR. Additionally, we address the challenges of scarcity and annotation bias in MLR datasets by introducing eight synthetic datasets (Ranked MNISTs) generated with varying significance-determining factors, providing an enriched and controllable experimental environment. We statistically demonstrate that our method accurately learns a representation of the positive rank order, which is consistent with the ground truth and proportional to the underlying significance values. Finally, we conduct comprehensive empirical experiments on both real-world and synthetic datasets, demonstrating the value of our proposed framework.

Busra Asan, Abdullah Akgül, Alper Unal et al.

Seasonal forecasting is a crucial task when it comes to detecting the extreme heat and colds that occur due to climate change. Confidence in the predictions should be reliable since a small increase in the temperatures in a year has a big impact on the world. Calibration of the neural networks provides a way to ensure our confidence in the predictions. However, calibrating regression models is an under-researched topic, especially in forecasters. We calibrate a UNet++ based architecture, which was shown to outperform physics-based models in temperature anomalies. We show that with a slight trade-off between prediction error and calibration error, it is possible to get more reliable and sharper forecasts. We believe that calibration should be an important part of safety-critical machine learning applications such as weather forecasters.

Gozdenur Demir, Asli Cekmis, Vahit Bugra Yesilkaynak et al.

Visual design is associated with the use of some basic design elements and principles. Those are applied by the designers in the various disciplines for aesthetic purposes, relying on an intuitive and subjective process. Thus, numerical analysis of design visuals and disclosure of the aesthetic value embedded in them are considered as hard. However, it has become possible with emerging artificial intelligence technologies. This research aims at a neural network model, which recognizes and classifies the design principles over different domains. The domains include artwork produced since the late 20th century; professional photos; and facade pictures of contemporary buildings. The data collection and curation processes, including the production of computationally-based synthetic dataset, is genuine. The proposed model learns from the knowledge of myriads of original designs, by capturing the underlying shared patterns. It is expected to consolidate design processes by providing an aesthetic evaluation of the visual compositions with objectivity.

Ufuk Demir, Atahan Ozer, Yusuf H. Sahin et al.

In recent years, deep learning based methods have shown success in essential medical image analysis tasks such as segmentation. Post-processing and refining the results of segmentation is a common practice to decrease the misclassifications originating from the segmentation network. In addition to widely used methods like Conditional Random Fields (CRFs) which focus on the structure of the segmented volume/area, a graph-based recent approach makes use of certain and uncertain points in a graph and refines the segmentation according to a small graph convolutional network (GCN). However, there are two drawbacks of the approach: most of the edges in the graph are assigned randomly and the GCN is trained independently from the segmentation network. To address these issues, we define a new neighbor-selection mechanism according to feature distances and combine the two networks in the training procedure. According to the experimental results on pancreas segmentation from Computed Tomography (CT) images, we demonstrate improvement in the quantitative measures. Also, examining the dynamic neighbors created by our method, edges between semantically similar image parts are observed. The proposed method also shows qualitative enhancements in the segmentation maps, as demonstrated in the visual results.

Melih Kandemir, Abdullah Akgül, Manuel Haussmann et al.

A probabilistic classifier with reliable predictive uncertainties i) fits successfully to the target domain data, ii) provides calibrated class probabilities in difficult regions of the target domain (e.g.\ class overlap), and iii) accurately identifies queries coming out of the target domain and rejects them. We introduce an original combination of Evidential Deep Learning, Neural Processes, and Neural Turing Machines capable of providing all three essential properties mentioned above for total uncertainty quantification. We observe our method on five classification tasks to be the only one that can excel all three aspects of total calibration with a single standalone predictor. Our unified solution delivers an implementation-friendly and compute efficient recipe for safety clearance and provides intellectual economy to an investigation of algorithmic roots of epistemic awareness in deep neural nets.

Alican Mertan, Damien Jade Duff, Gozde Unal

We review solutions to the problem of depth estimation, arguably the most important subtask in scene understanding. We focus on the single image depth estimation problem. Due to its properties, the single image depth estimation problem is currently best tackled with machine learning methods, most successfully with convolutional neural networks. We provide an overview of the field by examining key works. We examine non-deep learning approaches that mostly predate deep learning and utilize hand-crafted features and assumptions, and more recent works that mostly use deep learning techniques. The single image depth estimation problem is tackled first in a supervised fashion with absolute or relative depth information acquired from human or sensor-labeled data, or in an unsupervised way using unlabelled stereo images or video datasets. We also study multitask approaches that combine the depth estimation problem with related tasks such as semantic segmentation and surface normal estimation. Finally, we discuss investigations into the mechanisms, principles, and failure cases of contemporary solutions.

Yusuf H. Sahin, Alican Mertan, Gozde Unal

Learning new representations of 3D point clouds is an active research area in 3D vision, as the order-invariant point cloud structure still presents challenges to the design of neural network architectures. Recent works explored learning either global or local features or both for point clouds, however none of the earlier methods focused on capturing contextual shape information by analysing local orientation distribution of points. In this paper, we leverage on point orientation distributions around a point in order to obtain an expressive local neighborhood representation for point clouds. We achieve this by dividing the spherical neighborhood of a given point into predefined cone volumes, and statistics inside each volume are used as point features. In this way, a local patch can be represented by not only the selected point's nearest neighbors, but also considering a point density distribution defined along multiple orientations around the point. We are then able to construct an orientation distribution function (ODF) neural network that involves an ODFBlock which relies on mlp (multi-layer perceptron) layers. The new ODFNet model achieves state-of the-art accuracy for object classification on ModelNet40 and ScanObjectNN datasets, and segmentation on ShapeNet S3DIS datasets.

Alican Mertan, Yusuf Huseyin Sahin, Damien Jade Duff et al.

We propose a new approach for the problem of relative depth estimation from a single image. Instead of directly regressing over depth scores, we formulate the problem as estimation of a probability distribution over depth and aim to learn the parameters of the distributions which maximize the likelihood of the given data. To train our model, we propose a new ranking loss, Distributional Loss, which tries to increase the probability of farther pixel's depth being greater than the closer pixel's depth. Our proposed approach allows our model to output confidence in its estimation in the form of standard deviation of the distribution. We achieve state of the art results against a number of baselines while providing confidence in our estimations. Our analysis show that estimated confidence is actually a good indicator of accuracy. We investigate the usage of confidence information in a downstream task of metric depth estimation, to increase its performance.

Alican Mertan, Damien Jade Duff, Gozde Unal

We present a formulation of the relative depth estimation from a single image problem, as a ranking problem. By reformulating the problem this way, we were able to utilize literature on the ranking problem, and apply the existing knowledge to achieve better results. To this end, we have introduced a listwise ranking loss borrowed from ranking literature, weighted ListMLE, to the relative depth estimation problem. We have also brought a new metric which considers pixel depth ranking accuracy, on which our method is stronger.

Vahit Bugra Yesilkaynak, Yusuf H. Sahin, Gozde Unal

Deep neural network training without pre-trained weights and few data is shown to need more training iterations. It is also known that, deeper models are more successful than their shallow counterparts for semantic segmentation task. Thus, we introduce EfficientSeg architecture, a modified and scalable version of U-Net, which can be efficiently trained despite its depth. We evaluated EfficientSeg architecture on Minicity dataset and outperformed U-Net baseline score (40% mIoU) using the same parameter count (51.5% mIoU). Our most successful model obtained 58.1% mIoU score and got the fourth place in semantic segmentation track of ECCV 2020 VIPriors challenge.

Furkan Ozcelik, Ugur Alganci, Elif Sertel et al.

Convolutional Neural Networks (CNN)-based approaches have shown promising results in pansharpening of satellite images in recent years. However, they still exhibit limitations in producing high-quality pansharpening outputs. To that end, we propose a new self-supervised learning framework, where we treat pansharpening as a colorization problem, which brings an entirely novel perspective and solution to the problem compared to existing methods that base their solution solely on producing a super-resolution version of the multispectral image. Whereas CNN-based methods provide a reduced resolution panchromatic image as input to their model along with reduced resolution multispectral images, hence learn to increase their resolution together, we instead provide the grayscale transformed multispectral image as input, and train our model to learn the colorization of the grayscale input. We further address the fixed downscale ratio assumption during training, which does not generalize well to the full-resolution scenario. We introduce a noise injection into the training by randomly varying the downsampling ratios. Those two critical changes, along with the addition of adversarial training in the proposed PanColorization Generative Adversarial Networks (PanColorGAN) framework, help overcome the spatial detail loss and blur problems that are observed in CNN-based pansharpening. The proposed approach outperforms the previous CNN-based and traditional methods as demonstrated in our experiments.

Gulcin Baykal, Gozde Unal

Generative Adversarial Networks (GANs) triggered an increased interest in problem of image generation due to their improved output image quality and versatility for expansion towards new methods. Numerous GAN-based works attempt to improve generation by architectural and loss-based extensions. We argue that one of the crucial points to improve the GAN performance in terms of realism and similarity to the original data distribution is to be able to provide the model with a capability to learn the spatial structure in data. To that end, we propose the DeshuffleGAN to enhance the learning of the discriminator and the generator, via a self-supervision approach. Specifically, we introduce a deshuffling task that solves a puzzle of randomly shuffled image tiles, which in turn helps the DeshuffleGAN learn to increase its expressive capacity for spatial structure and realistic appearance. We provide experimental evidence for the performance improvement in generated images, compared to the baseline methods, which is consistently observed over two different datasets.

M. Jorge Cardoso, Aasa Feragen, Ben Glocker et al.

This compendium gathers all the accepted extended abstracts from the Second International Conference on Medical Imaging with Deep Learning (MIDL 2019), held in London, UK, 8-10 July 2019. Note that only accepted extended abstracts are listed here, the Proceedings of the MIDL 2019 Full Paper Track are published as Volume 102 of the Proceedings of Machine Learning Research (PMLR) http://proceedings.mlr.press/v102/.

Sahin Olut, Yusuf Huseyin Sahin, Ugur Demir et al.

Magnetic Resonance Angiography (MRA) has become an essential MR contrast for imaging and evaluation of vascular anatomy and related diseases. MRA acquisitions are typically ordered for vascular interventions, whereas in typical scenarios, MRA sequences can be absent in the patient scans. This motivates the need for a technique that generates inexistent MRA from existing MR multi-contrast, which could be a valuable tool in retrospective subject evaluations and imaging studies. In this paper, we present a generative adversarial network (GAN) based technique to generate MRA from T1-weighted and T2-weighted MRI images, for the first time to our knowledge. To better model the representation of vessels which the MRA inherently highlights, we design a loss term dedicated to a faithful reproduction of vascularities. To that end, we incorporate steerable filter responses of the generated and reference images inside a Huber function loss term. Extending the well- established generator-discriminator architecture based on the recent PatchGAN model with the addition of steerable filter loss, the proposed steerable GAN (sGAN) method is evaluated on the large public database IXI. Experimental results show that the sGAN outperforms the baseline GAN method in terms of an overlap score with similar PSNR values, while it leads to improved visual perceptual quality.

Ugur Demir, Gozde Unal

Area of image inpainting over relatively large missing regions recently advanced substantially through adaptation of dedicated deep neural networks. However, current network solutions still introduce undesired artifacts and noise to the repaired regions. We present an image inpainting method that is based on the celebrated generative adversarial network (GAN) framework. The proposed PGGAN method includes a discriminator network that combines a global GAN (G-GAN) architecture with a patchGAN approach. PGGAN first shares network layers between G-GAN and patchGAN, then splits paths to produce two adversarial losses that feed the generator network in order to capture both local continuity of image texture and pervasive global features in images. The proposed framework is evaluated extensively, and the results including comparison to recent state-of-the-art demonstrate that it achieves considerable improvements on both visual and quantitative evaluations.

Ugur Demir, Gozde Unal

Deep neural networks have shown promising results in image inpainting even if the missing area is relatively large. However, most of the existing inpainting networks introduce undesired artifacts and noise to the repaired regions. To solve this problem, we present a novel framework which consists of two stacked convolutional neural networks that inpaint the image and remove the artifacts, respectively. The first network considers the global structure of the damaged image and coarsely fills the blank area. Then the second network modifies the repaired image to cancel the noise introduced by the first network. The proposed framework splits the problem into two distinct partitions that can be optimized separately, therefore it can be applied to any inpainting algorithm by changing the first network. Second stage in our framework which aims at polishing the inpainted images can be treated as a denoising problem where a wide range of algorithms can be employed. Our results demonstrate that the proposed framework achieves significant improvement on both visual and quantitative evaluations.