Synthetic aperture radar (SAR) image simulators based on computer-aided drawing models play an important role in SAR applications, such as automatic target recognition and image interpretation. However, the accuracy of such simulators is due to geometric error and simplification in the electromagnetic calculation. In this letter, an end-to-end model was developed that could directly synthesize the desired images from the known image database. The model was based on generative adversarial nets (GANs), and its feasibility was validated by comparisons with real images and ray-tracing results. As a further step, the samples were synthesized at angles outside of the data set. However, the training process of GAN models was difficult, especially for SAR images which are usually affected by noise interference. The major failure modes were analyzed in experiments, and a clutter normalization method was proposed to ameliorate them. The results showed that the method improved the speed of convergence up to 10 times. The quality of the synthesized images was also improved.

Deep learning (DL) has been successfully applied to single image super-resolution (SISR), which aims at reconstructing a high-resolution (HR) image from its low-resolution (LR) counterpart. Different from most current DL-based methods, which perform reconstruction in the spatial domain, we use a scheme based in the frequency domain to reconstruct the HR image at various frequency bands. Further, we propose a method that incorporates the wavelet transform (WT) and the recursive Res-Net. The WT is applied to the LR image to divide it into various frequency components. Then, an elaborately designed network with recursive residual blocks is used to predict high-frequency components. Finally, the reconstructed image is obtained via the inverse WT. This paper has three main contributions: 1) an SISR scheme based on the frequency domain is proposed under a DL framework to fully exploit the potential to depict images at different frequency bands; 2) recursive block and residual learning in global and local manners are adopted to ease the training of the deep network, and the batch normalization layer is removed to increase the flexibility of the network, save memory, and promote speed; and 3) the low-frequency wavelet component is replaced by an LR image with more details to further improve performance. To validate the effectiveness of the proposed method, extensive experiments are performed using the NWPU-RESISC45 data set, and the results demonstrate that the proposed method outperforms several state-of-the-art methods in terms of both objective evaluation and subjective perspective.

High-resolution (HR) images are always preferred for many remote sensing applications, which can be obtained from their low-resolution (LR) counterparts via a technique referred to as super-resolution (SR). Among SR approaches, single image SR (SISR) methods aim at reconstructing the HR image from only one LR image. In this paper, a residual dense backprojection network (RDBPN)-based SISR method is proposed to promote the resolution of RGB remote sensing images with median- and large-scale factors. The proposed network consists of several residual dense backprojection blocks that contain two kinds of modules, named the upprojection module and the downprojection module, and these modules are densely connected in one block. Different from the chain-connected backprojection structure, the proposed method applies a residual backprojection block structure, which can utilize residual learning in both global and local manners. We further simplify the network by replacing the downprojection unit with the downscaling unit to accelerate the speed of reconstruction, and this implementation is called fast RDBPN (FRDBPN). Several experiments under the UC Merced data set are conducted to validate the effectiveness of the proposed method, and the results indicate that: 1) the proposed residual block structure is superior to the chain-connected structure; 2) FRDBPN achieves a speedup of about 1.3 times with similar and even better-reconstructed performance in comparison with RDBPN; and 3) RDBPN and FRDBPN outperform several state-of-the-art methods in terms of both quantitative evaluation and visual quality.

Text-to-image (T2I) research has grown explosively in the past year, owing to the large-scale pre-trained diffusion models and many emerging personalization and editing approaches. Yet, one pain point persists: the text prompt engineering, and searching high-quality text prompts for customized results is more art than science. Moreover, as commonly argued: “an image is worth a thousand words” - the attempt to describe a desired image with texts often ends up being ambiguous and cannot comprehensively cover delicate visual details, hence necessitating more additional controls from the visual domain. In this paper, we take a bold step forward: taking “Text” out of a pretrained T2I diffusion model, to reduce the burdensome prompt engineering efforts for users. Our proposed frame-work, Prompt-Free Diffusion, relies on only visual inputs to generate new images: it takes a reference image as “context”, an optional image structural conditioning, and an initial noise, with absolutely no text prompt. The core architecture behind the scene is Semantic Context Encoder (SeeCoder), substituting the commonly used CLIP-based or LLM-based text encoder. The reusability of SeeCoder also makes it a convenient drop-in component: one can also pre-train a SeeCoder in one T2I model and reuse it for another. Through extensive experiments, Prompt-Free Diffusion is experimentally found to (i) outperform prior exemplar-based image synthesis approaches; (ii) perform on par with state-of-the-art T2I models using prompts following the best practice; and (iii) be naturally extensible to other downstream applications such as anime figure generation and virtual try-on, with promising quality. Our code and models will be open-sourced.

In this paper a non-parametric model based on Wasserstein CNN is proposed for color correction. It is suitable for large-scale remote sensing image preprocessing from multiple sources under various viewing conditions, including illumination variances, atmosphere disturbances, and sensor and aspect angles. Color correction aims to alter the color palette of an input image to a standard reference which does not suffer from the mentioned disturbances. Most of current methods highly depend on the similarity between the inputs and the references, with respect to both the contents and the conditions, such as illumination and atmosphere condition. Segmentation is usually necessary to alleviate the color leakage effect on the edges. Different from the previous studies, the proposed method matches the color distribution of the input dataset with the references in a probabilistic optimal transportation framework. Multi-scale features are extracted from the intermediate layers of the lightweight CNN model and are utilized to infer the undisturbed distribution. The Wasserstein distance is utilized to calculate the cost function to measure the discrepancy between two color distributions. The advantage of the method is that no registration or segmentation processes are needed, benefiting from the local texture processing potential of the CNN models. Experimental results demonstrate that the proposed method is effective when the input and reference images are of different sources, resolutions, and under different illumination and atmosphere conditions.

Synthetic aperture radar tomography (TomoSAR) is an important 3D mapping method. Traditional TomoSAR requires a large number of observation orbits however, it is hard to meet the requirement of massive orbits. While on the one hand, this is due to funding constraints, on the other hand, because the target scene is changing over time and each observation orbit consumes lots of time, the number of orbits can be fewer as required within a narrow time window. When the number of observation orbits is insufficient, the signal-to-noise ratio (SNR), peak-to-sidelobe ratio (PSR), and resolution of 3D reconstruction results will decline severely, which seriously limits the practical application of TomoSAR. In order to solve this problem, we propose to use a deep learning network to improve the resolution and SNR of 3D reconstruction results under the condition of very few observation orbits by learning the prior distribution of targets. We use all available orbits to reconstruct a high resolution target, while only very few (around 3) orbits to reconstruct a low resolution input. The low-res and high-res 3D voxel-grid pairs are used to train a 3D super-resolution (SR) CNN (convolutional neural network) model, just like ordinary 2D image SR tasks. Experiments on the Civilian Vehicle Radar dataset show that the proposed deep learning algorithm can effectively improve the reconstruction both in quality and in quantity. In addition, the model also shows good generalization performance for targets not shown in the training set.