Photographs taken through a glass window are susceptible to disturbances due to reflection. Therefore, single image reflection removal is crucial to image quality enhancement. In this paper, a novel learning architecture that can address this ill-posed problem is proposed. First, a novel reflection removal pipeline was designed to reconstruct the missing information caused by the camera imaging process using the proposed missing recovery network. Second, to address the issues in existing reflection removal strategies, we revisit several auxiliary priors and integrate them by defining an energy function. To solve the energy function, a convolutional neural network-based optimization scheme was proposed. Finally, we investigated the dark channel responses of reflection and clean images and found an interesting way to distinguish between these two types of images. We prove this property mathematically and propose a novel loss function called dark channel loss to improve performance. Experiments show that the proposed method outperforms state-of-the-art reflection removal methods both quantitatively and qualitatively.

In image processing, smoke may degrade visibility and deteriorate the performance of high-level vision applications. Therefore, single image smoke removal is crucial for computer vision. Currently, existing smoke removal algorithms mainly leverage handcrafted priors. Moreover, these methods usually apply haze removal methods to perform smoke removal due to the similarity between smoke and haze. However, these methods cannot sufficiently address the degradation of thick smoke and may suffer from residual smoke and color distortion problems due to the non-global and non-homogeneous distribution of smoke. In this paper, to solve the aforementioned problems, an end-to-end deep neural network called DesmokeNet is proposed. We construct a two-stage recovered pipeline to remove the smoke in different thicknesses. The light and thick smoke is first removed locally by the smoke removal network (SRN). The missing pixels in the thick smoke are then recovered by the pixel compensation network (PCN). Moreover, we proposed the thickness-aware pixel loss and the dark channel loss to suppress the residual smoke. To further increase the discriminative ability of the DesmokeNet, we proposed self-attentive feature consensus loss and multi-level contrastive regularization loss to improve the performance of smoke removal. Finally, to train the proposed method, we construct the first large-scale dataset containing synthetic and real-world data. Extensive experiments show that the proposed method outperforms favorably against other state-of-the-art methods quantitatively and qualitatively.

Snow is a highly complicated atmospheric phenomenon that usually contains snowflake, snow streak, and veiling effect (similar to the haze or the mist). In this literature, we propose a single image desnowing algorithm to address the diversity of snow particles in shape and size. First, to better represent the complex snow shape, we apply the dual-tree wavelet transform and propose a complex wavelet loss in the network. Second, we propose a hierarchical decomposition paradigm in our network for better under-standing the different sizes of snow particles. Last, we propose a novel feature called the contradict channel (CC) for the snow scenes. We find that the regions containing the snow particles tend to have higher intensity in the CC than that in the snow-free regions. We leverage this discriminative feature to construct the contradict channel loss for improving the performance of snow removal. Moreover, due to the limitation of existing snow datasets, to simulate the snow scenarios comprehensively, we propose a large-scale dataset called Comprehensive Snow Dataset (CSD). Experimental results show that the proposed method can favorably outperform existing methods in three synthetic datasets and real-world datasets. The code and dataset are released in https://github.com/weitingchen83/ICCV2021-Single-Image-Desnowing-HDCWNet.

Single image haze removal is crucial in computer vision. In open literatures, two kinds of dehazing strategies (prior-based and learning-based methods) have been developed. However, they have a trade-off between detail preservation and the image quality. Prior-based methods reconstruct the detail well but have lower image quality while learning-based methods achieve better recovered quality but lose the detail. In this paper, to mitigate this dilemma, a hierarchical architecture using the discrete wavelet transform (DWT) is proposed. It divides the dehazing problem into two parts: detail and background reconstruction. Based on investigating how haze affects the image in the wavelet domain, two networks for detail and background reconstruction are proposed. To avoid color distortion and the detail loss, the anti-vanish wavelet loss and the bound penalty are proposed. The multi-level wavelet component discriminator is proposed for further improvement. Experiments show that the proposed network can achieve superior performance in all metrics.

Images captured in a hazy environment usually suffer from bad visibility and missing information. Over many years, learning-based and handcrafted prior-based dehazing algorithms have been rigorously developed. However, both algorithms exhibit some weaknesses in terms of haze removal performance. Therefore, in this work, we have proposed the patch-map-based hybrid learning DehazeNet, which integrates these two strategies by using a hybrid learning technique involving the patch map and a bi-attentive generative adversarial network. In this method, the reasons limiting the performance of the dark channel prior (DCP) have been analyzed. A new feature called the patch map has been defined for selecting the patch size adaptively. Using this map, the limitations of the DCP (e.g., color distortion and failure to recover images involving white scenes) can be addressed efficiently. In addition, to further enhance the performance of the method for haze removal, a patch-map-based DCP has been embedded into the network, and this module has been trained with the atmospheric light generator, patch map selection module, and refined module simultaneously. A combination of traditional and learning-based methods can efficiently improve the haze removal performance of the network. Experimental results show that the proposed method can achieve better reconstruction results compared to other state-of-the-art haze removal algorithms.