Visual foundation models (VFMs) have demonstrated impressive generalization capabilities in computer vision tasks. Previous studies show that fine-tuning VFMs with learnable tokens can achieve better generalization performance than full-parameter fine-tuning. The problem we need to address is how to learn the tokens that focus on the content information while ignoring the influence of style. For this purpose, we propose a novel Dual-Branch Content-enhanced Token (DBCT) learning framework. Specifically, we construct a style-suppressing branch, which contains a Style-sensitive Channel Suppression (SCS) module to transform the frozen VFM features into style-suppressed features, enabling the learning of style-invariant tokens. In addition, to compensate for the content degradation caused by the style-suppressing branch, we introduce a content-preserving branch that directly takes the frozen VFM features as input to learn content-focused tokens. Meanwhile, we propose a Token-query Linking (TLink) strategy to connect the two sets of tokens with the queries in the decoder. Through extensive experiments, our method achieves advanced results on various benchmarks.

Images captured under low light conditions are often affected by intense noise, which may become more pronounced during image enhancement, resulting in poor visual quality. The aim of this paper is to establish an effective low-light image enhancement model that can suppress noise and artifacts while preserving image details. To deal with intense noise, we propose a Weighted Low-Rank Tensor regularization Retinex (WLRT-Retinex) model, which introduces weighted low-rank tensor priors in the Retinex decomposition process to suppress noise and artifacts in the reflectance. Furthermore, since noise in dark areas is typically more severe, we introduce an illumination-aware weighting scheme in the total variation regularization term of the reflectance, which helps achieve adaptive denoising and preserve details in bright areas. Experiments on seven challenging datasets demonstrate the effectiveness of the proposed method, achieving better or comparable performance compared with state-of-the-art methods. Our code is available at https://github.com/YangWeipengscut/WLRT-Retinex.

Recently, deep neural networks have been crucial techniques for image salient detection. However, two difficulties prevent the development of deep learning in video saliency detection. The first one is that the traditional static network cannot conduct a robust motion estimation in videos. The other is that the data-driven deep learning is in lack of sufficient manually annotated pixel-wise ground truths for video saliency network training. In this paper, we propose a multi-scale spatiotemporal convolutional LSTM network (MSST-ConvLSTM) to incorporate spatial and temporal cues for video salient objects detection. Furthermore, as manually pixel-wised labeling is very time-consuming, we sign lots of coarse labels, which are mixed with fine labels to train a robust saliency prediction model. Experiments on the widely used challenging benchmark datasets (e.g., FBMS and DAVIS) demonstrate that the proposed approach has competitive performance of video saliency detection compared with the state-of-the-art saliency models.