Showing papers by "Xin Jin published in 2018"

Unsupervised Single Image Deraining with Self-supervised Constraints

Abstract: Most existing single image deraining methods require learning supervised models from a large set of paired synthetic training data, which limits their generality, scalability and practicality in real-world multimedia applications. Besides, due to lack of labeled-supervised constraints, directly applying existing unsupervised frameworks to the image deraining task will suffer from low-quality recovery. Therefore, we propose an Unsupervised Deraining Generative Adversarial Network (UD-GAN) to tackle above problems by introducing self-supervised constraints from the intrinsic statistics of unpaired rainy and clean images. Specifically, we firstly design two collaboratively optimized modules, namely Rain Guidance Module (RGM) and Background Guidance Module (BGM), to take full advantage of rainy image characteristics: The RGM is designed to discriminate real rainy images from fake rainy images which are created based on outputs of the generator with BGM. Simultaneously, the BGM exploits a hierarchical Gaussian-Blur gradient error to ensure background consistency between rainy input and de-rained output. Secondly, a novel luminance-adjusting adversarial loss is integrated into the clean image discriminator considering the built-in luminance difference between real clean images and derained images. Comprehensive experiment results on various benchmarking datasets and different training settings show that UD-GAN outperforms existing image deraining methods in both quantitative and qualitative comparisons.

A Decomposed Dual-Cross Generative Adversarial Network for Image Rain Removal.

Abstract: Rain removal is important for many computer vision applications, such as surveillance, autonomous car, etc. Traditionally, rain removal is regarded as a signal removal problem which usually causes over-smoothing by removing texture details in non-rain background regions. This paper considers the issue of rain removal from a completely different perspective, to treat rain removal as a signal decomposition problem. Specifically, we decompose the rain image into two components, namely non-rain background image and rain streaks image. Then, we introduce an adversarial training mechanism to synthesize non-rain background image and rain streaks image in a Dual-Cross manner, which makes the two adversarial branches interact with each other, archiving a win-win result ultimately. The proposed Decomposed Dual-Cross Generative Adversarial Network (DDC-GAN) shows significantly performance improvement compared with stateof-the-art methods on both synthetic and real-world images in terms of qualitative and quantitative measures (over 3dB gains in PSNR).

An Intelligent Scheduling Approach for Electric Power Generation

Abstract: Although China is vigorously developing clean energy and nuclear power, the thermal power generation (mainly coal power) is still the most important power generation method at present. Economic load dispatch (ELD) is a typical optimization problem in power systems which lots of researchers are trying to explore. The purpose of ELD is to increase the efficiency of thermal power generation under the conditions of load and operational constraints. When it comes to power generation scheduling, manual operation is still the main form, which is inefficient. In order to use a large amount of historical power generation data to improve the efficiency of power generation scheduling and achieve the effect of energy conservation, we propose an intelligent power generation scheduling system based on Deep neural networks (DNN) and Ant colony optimization (ACO). Experiments show that our DNN algorithm can predict the unit coal consumption precisely. Compared with the dynamic programming algorithm and equal differential increment rate algorithm, ACO can complete power generation scheduling tasks more quickly and efficiently.

Augmented Coarse-to-Fine Video Frame Synthesis with Semantic Loss

Abstract: Existing video frame synthesis works suffer from improving perceptual quality and preserving semantic representation ability. In this paper, we propose a Progressive Motion-texture Synthesis Network (PMSN) to address this problem. Instead of learning synthesis from scratch, we introduce augmented inputs to compensate texture details and motion information. Specifically, a coarse-to-fine guidance scheme with a well-designed semantic loss is presented to improve the capability of video frame synthesis. As shown in the experiments, our proposed PMSN promises excellent quantitative results, visual effects, and generalization ability compared with traditional solutions.

Multiscale Progressive Image Compression Network Guided by Learnable Just Noticeable Distortion

Abstract: One key challenge to the learning-based image compression is that adaptive bit allocation is crucial for compression effectiveness but can hardly be trained into a neural network. Hereby, in this work, We presents an end-to-end trainable image compression framework, named Multi-scale Progressive Network (MPN) to achieve spatially variant bit allocation and rate control through the guidance of a novel learnable just noticeable distortion (JND) map. Specifically, MPN’s encoder archives multi-scale feature representation through a three-branched structure. Each branch employs an independent feature extraction strategy for the specific receptive field and merge progressively under the guidance of corresponding learnable JND maps that generated by our proposed Bit-Allocation sub-Network (BAN), which make MPN focus on the areas where attract the human visual system (HVS) and preserve more texture of the image during the compression procedure. Finally, a hybrid objective function is introduced to further make MPN more efficient and mimic the discriminative characteristics of the human visual system (HVS). Experiments show that MPN significantly outperforms traditional JPEG, JPEG 2000 and few state-of-art learning-based methods by multi-scale structural similarity (MS-SSIM) index, and has the ability to produce the much better visual result with rich textures, sharp edges, and fewer artifacts.