Recently, learning-based Underwater Image Enhancement (UIE) methods have demonstrated promising performance. However, existing learning-based methods still face two challenges. 1) They rarely consider the inconsistent degradation levels in different spatial regions and spectral bands simultaneously. 2) They treat all regions equally, ignoring that the regions with high-frequency details are more difficult to reconstruct. To address these challenges, we propose a novel UIE method based on spatial-spectral dual-domain adaptive learning, termed SS-UIE. Specifically, we first introduce a spatial-wise Multi-scale Cycle Selective Scan (MCSS) module and a Spectral-Wise Self-Attention (SWSA) module, both with linear complexity, and combine them in parallel to form a basic Spatial-Spectral block (SS-block). Benefiting from the global receptive field of MCSS and SWSA, SS-block can effectively model the degradation levels of different spatial regions and spectral bands, thereby enabling degradation level-based dual-domain adaptive UIE. By stacking multiple SS-blocks, we build our SS-UIE network. Additionally, a Frequency-Wise Loss (FWL) is introduced to narrow the frequency-wise discrepancy and reinforce the model's attention on the regions with high-frequency details. Extensive experiments validate that the SS-UIE technique outperforms state-of-the-art UIE methods while requiring cheaper computational and memory costs.

Recently, learning-based hyperspectral anomaly detection (HAD) methods have demonstrated outstanding performance, dominating mainstream research. However, the existing learning-based approaches still have two issues: 1) they rarely consider both the spatial sparsity and the interspectral similarity of hyperspectral imagery (HSI) simultaneously and 2) they treat all regions equally, often overlooking the importance of high-frequency information in HSI, which is key to distinguish background and anomalies. To address these challenges, we propose a novel HAD method based on spatial-spectral adaptive dual-domain learning, termed SSHAD. Specifically, we first introduce the spatial-wise selected state space module (SSSM) with linear complexity and the spectral-wise frequency division self-attention module (FDSM), which are combined in parallel to form a spatial-spectral block (SS-block). The SSSM captures the global receptive field by scanning the HSI spatial dimension through a multidirectional scanning mechanism. The FDSM extracts high-frequency and low-frequency information from the HSI via the discrete wavelet transform (DWT) and applies multiscale convolution and self-similarity attention respectively, ensuring the suppression of anomalies during the reconstruction process. This parallel structure enables the network to model cross-window connections, expanding its receptive field while maintaining linear complexity. We use the SS-block as the main component of our adaptive dual-domain learning network, forming SSHAD. Furthermore, we introduce a frequency-wise loss function to inhibit the reconstruction of high-frequency anomalies during background reconstruction. Comprehensive experiments conducted on four public datasets and two unmanned aerial vehicle (UAV)-borne datasets validate the superiority and effectiveness of SSHAD. The code will be publicly available athttps://github.com/CZhu0066/SSHAD.

Accurate oral disease segmentation is a challenging task, for three major reasons: 1) The same type of oral disease has a diversity of size, color and texture; 2) The boundary between oral lesions and their surrounding mucosa is not sharp; 3) There is a lack of public large-scale oral disease segmentation datasets. To address these issues, we first report an oral disease segmentation network termed Oralformer, which enables to tackle multiple oral diseases. Specifically, we use a parallel design to combine local-window self-attention (LWSA) with channel-wise convolution (CWC), modeling cross-window connections to enlarge the receptive fields while maintaining linear complexity. Meanwhile, we connect these two branches with bi-directional interactions to form a basic parallel Transformer block namely LC-block. We insert the LC-block as the main building block in a U-shape encoder-decoder architecture to form Oralformer. Second, we introduce an uncertainty-driven self-adaptive loss function which can reinforce the network's attention on the lesion's edge regions that are easily confused, thus improving the segmentation accuracy of these regions. Third, we construct a large-scale oral disease segmentation (ODS) dataset containing 2602 image pairs. It covers three common oral diseases (including dental plaque, calculus and caries) and all age groups, which we hope will advance the field. Extensive experiments on six challenging datasets show that our Oralformer achieves state-of-the-art segmentation accuracy, and presents advantages in terms of generalizability and real-time segmentation efficiency (35fps). The code and ODS dataset will be publicly available at https://github.com/LintaoPeng/Oralformer.

The light absorption and scattering of underwater impurities lead to poor underwater imaging quality. The existing data-driven based underwater image enhancement (UIE) techniques suffer from the lack of a large-scale dataset containing various underwater scenes and high-fidelity reference images. Besides, the inconsistent attenuation in different color channels and space areas is not fully considered for boosted enhancement. In this work, we built a large scale underwater image (LSUI) dataset, which covers more abundant underwater scenes and better visual quality reference images than existing underwater datasets. The dataset contains 4279 real-world underwater image groups, in which each raw image's clear reference images, semantic segmentation map and medium transmission map are paired correspondingly. We also reported an U-shape Transformer network where the transformer model is for the first time introduced to the UIE task. The U-shape Transformer is integrated with a channel-wise multi-scale feature fusion transformer (CMSFFT) module and a spatial-wise global feature modeling transformer (SGFMT) module specially designed for UIE task, which reinforce the network's attention to the color channels and space areas with more serious attenuation. Meanwhile, in order to further improve the contrast and saturation, a novel loss function combining RGB, LAB and LCH color spaces is designed following the human vision principle. The extensive experiments on available datasets validate the state-of-the-art performance of the reported technique with more than 2dB superiority. The dataset and demo code are available at https://bianlab.github.io/.