Multi-exposure fusion is a high dynamic range (HDR) imaging technique that combines multiple low dynamic range (LDR) images of a scene with varying exposure times to produce a single high-quality HDR image. Since each LDR frame is captured with a different exposure time (bias), it is crucial to extract meaningful features from each differently-exposed LDR frame for producing high-quality HDR images. This paper introduces a new contrastive learning method that provides a versatile way of extracting characteristic features from LDR frames by considering the relationship between LDR frames. Additionally, we introduce Representative Feature-Guided Transformer (RFGHDR), a new architecture that utilizes contrastive-learned representations to improve frame alignment and merging. Based on extensive experiments on various datasets, we have found that the RFG-HDR performs better than existing multi-exposure HDR imaging methods in terms of various evaluation metrics. Our work will be released on https://github.com/KeuntekLee/RFG-HDR.

There have been many image denoisers using deep neural networks, which outperform conventional model-based methods by large margins. Recently, self-supervised methods have attracted attention because constructing a large real noise dataset for supervised training is an enormous burden. The most representative self-supervised denoisers are based on blind-spot networks, which exclude the receptive field’s center pixel. However, excluding any input pixel is abandoning some information, especially when the input pixel at the corresponding output position is excluded. In addition, a standard blind-spot network fails to reduce real camera noise due to the pixel-wise correlation of noise, though it successfully removes independently distributed synthetic noise. Hence, to realize a more practical denoiser, we propose a novel self-supervised training framework that can remove real noise. For this, we derive the theoretic upper bound of a supervised loss where the network is guided by the downsampled blinded output. Also, we design a conditional blind-spot network (C-BSN), which selectively controls the blindness of the network to use the center pixel information. Furthermore, we exploit a random subsampler to decorrelate noise spatially, making the C-BSN free of visual artifacts that were often seen in downsample-based methods. Extensive experiments show that the proposed C-BSN achieves state-of-the-art performance on real-world datasets as a self-supervised denoiser and shows qualitatively pleasing results without any post-processing or refinement.

Real-noise denoising is a challenging task because the statistics of real-noise do not follow the normal distribution, and they are also spatially and temporally changing. In order to cope with various and complex real-noise, we propose a well-generalized denoising architecture and a transfer learning scheme. Specifically, we adopt an adaptive instance normalization to build a denoiser, which can regularize the feature map and prevent the network from overfitting to the training set. We also introduce a transfer learning scheme that transfers knowledge learned from synthetic-noise data to the real-noise denoiser. From the proposed transfer learning, the synthetic-noise denoiser can learn general features from various synthetic-noise data, and the real-noise denoiser can learn the real-noise characteristics from real data. From the experiments, we find that the proposed denoising method has great generalization ability, such that our network trained with synthetic-noise achieves the best performance for Darmstadt Noise Dataset (DND) among the methods from published papers. We can also see that the proposed transfer learning scheme robustly works for real-noise images through the learning with a very small number of labeled data.

Recently, many convolutional neural networks for single image super-resolution (SISR) have been proposed, which focus on reconstructing the high-resolution images in terms of objective distortion measures. However, the networks trained with objective loss functions generally fail to reconstruct the realistic fine textures and details that are essential for better perceptual quality. Recovering the realistic details remains a challenging problem, and only a few works have been proposed which aim at increasing the perceptual quality by generating enhanced textures. However, the generated fake details often make undesirable artifacts and the overall image looks somewhat unnatural. Therefore, in this paper, we present a new approach to reconstructing realistic super-resolved images with high perceptual quality, while maintaining the naturalness of the result. In particular, we focus on the domain prior properties of SISR problem. Specifically, we define the naturalness prior in the low-level domain and constrain the output image in the natural manifold, which eventually generates more natural and realistic images. Our results show better naturalness compared to the recent super-resolution algorithms including perception-oriented ones.

This paper presents a hierarchical convolutional neural network (CNN) for single image super-resolution (SISR), which exploits the controlled multi-context features. We focus on the method to extract more enriched features than the case of using fixed size kernels. For this, we attempt to bring out the best of the given parameter capacity through the design of some sophisticated networks in a hierarchical manner. First, we exploit the multi-kernel dilated convolution for extracting multi-size contexts from the image and combine them with the proposed trainable parameters. The multi-kernel network with some new pre-and post-processing blocks forms our basic building block. Then the basic building blocks are densely connected with a new feature fusion schemes, which makes the upper level building block. Then, by connecting the upper level blocks, we can use various features which can enrich the representation of the images. In the experiments, it is shown that the proposed method achieves significant PSNR gain compared to recent lightweight models with comparable numbers of parameters.

This paper presents a co-saliency detection algorithm based on clustering and diffusion process. For each image in a set, intra saliency maps are constructed from the measure of boundary and contrast priors. Then, segmented regions of all images are clustered based on features of colors, intra saliency and coherence of saliency. Co-saliency of each cluster is computed from combination of foreground probability and coherence of the cluster. The co-saliency of cluster is propagated over the segmented regions according to affinity between the cluster and segments. In addition, we adopt an intra image diffusion process from a graph with learned fully affinity in order to improve spatial consistency of co-saliency maps. Experimental results show that our algorithm yields better results compared to the state-of-the-art methods in terms of precision-recall curve, visual plausibility and computational cost.