Search

Many deep learning methods for image inpainting rely on the encoder-decoder architecture to estimate missing contents. When guidance information from uncorrupted regions cannot be adequately represented or utilized, the encoder may have difficulty handling the rich surrounding or background pixels, and the decoder could not recover visually sophisticated or realistic content. This paper proposes an effective multi-scale optimization network to alleviate these issues and generate coherent results with fine details. It adaptively encodes multi-receptive fields feature maps and puts multi-scale outputs into a discriminator to guide training. Specifically, we propose a Multi-Receptive feature maps & masks Selective Fusion (MRSF) operator that can adaptively extract features in different receptive fields to handle sophisticated destroyed images. Then a multi-gradient discriminator (MGD) module uses the intermediate features of the discriminator to guide the generator to produce results with natural textures and semantically real contents. Experiments on several benchmark datasets demonstrate that the proposed method can synthesize more realistic and coherent image content.

The reconstruction of missing pixels is essential for remote sensing images, as they often suffer from problems such as covering, dead pixels, and scan line corrector (SLC)-off. Image inpainting techniques can solve these problems, as they can generate realistic content for the unknown regions of an image based on the known regions. Recently, convolutional neural network (CNN)-based inpainting methods have integrated the attention mechanism to improve inpainting performance, as they can capture long-range dependencies and adapt to inputs in a flexible manner. However, to obtain the attention map for each feature, they compute the similarities between the feature and the entire feature map, which may introduce noise from irrelevant features. To address this problem, we propose a novel adaptive attention (Ada-attention) that uses an offset position subnet to adaptively select the most relevant keys and values based on self-attention. This enables the attention to be focused on essential features and model more informative dependencies on the global range. Ada-attention first employs an offset subnet to predict offset position maps on the query feature map; then, it samples the most relevant features from the input feature map based on the offset position; next, it computes key and value maps for self-attention using the sampled features; finally, using the query, key and value maps, the self-attention outputs the reconstructed feature map. Based on Ada-attention, we customized a u-shaped adaptive-attention completing network (AACNet) to reconstruct missing regions. Experimental results on several digital remote sensing and natural image datasets, using two image inpainting models and two remote sensing image reconstruction approaches, demonstrate that the proposed AACNet achieves a good quantitative performance and good visual restoration results with regard to object integrity, texture/edge detail, and structural consistency. Ablation studies indicate that Ada-attention outperforms self-attention in terms of PSNR by 0.66%, SSIM by 0.74%, and MAE by 3.9%, and can focus on valuable global features using the adaptive offset subnet. Additionally, our approach has also been successfully applied to remove real clouds in remote sensing images, generating credible content for cloudy regions.

Due to sensor malfunctions and poor atmospheric conditions, remote sensing images often miss important information/pixels, which affects downstream tasks, therefore requiring reconstruction. Current image reconstruction methods use deep convolutional neural networks to improve inpainting performances as they have a powerful modeling capability. However, deep convolutional networks learn different features with the same group of convolutional kernels, which restricts their ability to handle diverse image corruptions and often results in color discrepancy and blurriness in the recovered images. To mitigate this problem, in this paper, we propose an operator called Bilateral Convolution (BC) to adaptively preserve and propagate information from known regions to missing data regions. On the basis of vanilla convolution, the BC dynamically propagates more confident features, which weights the input features of a patch according to their spatial location and feature value. Furthermore, to capture different range dependencies, we designed a Multi-range Window Attention (MWA) module, in which the input feature is divided into multiple sizes of non-overlapped patches for several heads, and then these feature patches are processed by the window self-attention. With BC and MWA, we designed a bilateral convolution network for image inpainting. We conducted experiments on remote sensing datasets and several typical image inpainting datasets to verify the effectiveness and generalization of our network. The results show that our network adaptively captures features between known and unknown regions, generates appropriate content for various corrupted images, and has a competitive performance compared with state-of-the-art methods.