Jun Chen

Yawei Li, Kai Zhang, Radu Timofte et al. · eth-zurich, tencent-ai

This paper reviews the NTIRE 2022 challenge on efficient single image super-resolution with focus on the proposed solutions and results. The task of the challenge was to super-resolve an input image with a magnification factor of $\times$4 based on pairs of low and corresponding high resolution images. The aim was to design a network for single image super-resolution that achieved improvement of efficiency measured according to several metrics including runtime, parameters, FLOPs, activations, and memory consumption while at least maintaining the PSNR of 29.00dB on DIV2K validation set. IMDN is set as the baseline for efficiency measurement. The challenge had 3 tracks including the main track (runtime), sub-track one (model complexity), and sub-track two (overall performance). In the main track, the practical runtime performance of the submissions was evaluated. The rank of the teams were determined directly by the absolute value of the average runtime on the validation set and test set. In sub-track one, the number of parameters and FLOPs were considered. And the individual rankings of the two metrics were summed up to determine a final ranking in this track. In sub-track two, all of the five metrics mentioned in the description of the challenge including runtime, parameter count, FLOPs, activations, and memory consumption were considered. Similar to sub-track one, the rankings of five metrics were summed up to determine a final ranking. The challenge had 303 registered participants, and 43 teams made valid submissions. They gauge the state-of-the-art in efficient single image super-resolution.

Shirui Huang, Keyan Wang, Huan Liu et al.

Despite the remarkable achievement of recent underwater image restoration techniques, the lack of labeled data has become a major hurdle for further progress. In this work, we propose a mean-teacher based Semi-supervised Underwater Image Restoration (Semi-UIR) framework to incorporate the unlabeled data into network training. However, the naive mean-teacher method suffers from two main problems: (1) The consistency loss used in training might become ineffective when the teacher's prediction is wrong. (2) Using L1 distance may cause the network to overfit wrong labels, resulting in confirmation bias. To address the above problems, we first introduce a reliable bank to store the "best-ever" outputs as pseudo ground truth. To assess the quality of outputs, we conduct an empirical analysis based on the monotonicity property to select the most trustworthy NR-IQA method. Besides, in view of the confirmation bias problem, we incorporate contrastive regularization to prevent the overfitting on wrong labels. Experimental results on both full-reference and non-reference underwater benchmarks demonstrate that our algorithm has obvious improvement over SOTA methods quantitatively and qualitatively. Code has been released at https://github.com/Huang-ShiRui/Semi-UIR.

Jun Chen, Haishan Ye, Mengmeng Wang et al.

The conjugate gradient method is a crucial first-order optimization method that generally converges faster than the steepest descent method, and its computational cost is much lower than that of second-order methods. However, while various types of conjugate gradient methods have been studied in Euclidean spaces and on Riemannian manifolds, there is little study for those in distributed scenarios. This paper proposes a decentralized Riemannian conjugate gradient descent (DRCGD) method that aims at minimizing a global function over the Stiefel manifold. The optimization problem is distributed among a network of agents, where each agent is associated with a local function, and the communication between agents occurs over an undirected connected graph. Since the Stiefel manifold is a non-convex set, a global function is represented as a finite sum of possibly non-convex (but smooth) local functions. The proposed method is free from expensive Riemannian geometric operations such as retractions, exponential maps, and vector transports, thereby reducing the computational complexity required by each agent. To the best of our knowledge, DRCGD is the first decentralized Riemannian conjugate gradient algorithm to achieve global convergence over the Stiefel manifold.

Kangdi Shi, Muhammad Alrabeiah, Jun Chen

The advent of deep learning in the past decade has significantly helped advance image inpainting. Although achieving promising performance, deep learning-based inpainting algorithms still struggle from the distortion caused by the fusion of structural and contextual features, which are commonly obtained from, respectively, deep and shallow layers of a convolutional encoder. Motivated by this observation, we propose a novel progressive inpainting network that maintains the structural and contextual integrity of a processed image. More specifically, inspired by the Gaussian and Laplacian pyramids, the core of the proposed network is a feature extraction module named GLE. Stacking GLE modules enables the network to extract image features from different image frequency components. This ability is important to maintain structural and contextual integrity, for high frequency components correspond to structural information while low frequency components correspond to contextual information. The proposed network utilizes the GLE features to progressively fill in missing regions in a corrupted image in an iterative manner. Our benchmarking experiments demonstrate that the proposed method achieves clear improvement in performance over many state-of-the-art inpainting algorithms.

Yangyi Liu, Huan Liu, Liangyan Li et al.

Recent years have witnessed an increased interest in image dehazing. Many deep learning methods have been proposed to tackle this challenge, and have made significant accomplishments dealing with homogeneous haze. However, these solutions cannot maintain comparable performance when they are applied to images with non-homogeneous haze, e.g., NH-HAZE23 dataset introduced by NTIRE challenges. One of the reasons for such failures is that non-homogeneous haze does not obey one of the assumptions that is required for modeling homogeneous haze. In addition, a large number of pairs of non-homogeneous hazy image and the clean counterpart is required using traditional end-to-end training approaches, while NH-HAZE23 dataset is of limited quantities. Although it is possible to augment the NH-HAZE23 dataset by leveraging other non-homogeneous dehazing datasets, we observe that it is necessary to design a proper data-preprocessing approach that reduces the distribution gaps between the target dataset and the augmented one. This finding indeed aligns with the essence of data-centric AI. With a novel network architecture and a principled data-preprocessing approach that systematically enhances data quality, we present an innovative dehazing method. Specifically, we apply RGB-channel-wise transformations on the augmented datasets, and incorporate the state-of-the-art transformers as the backbone in the two-branch framework. We conduct extensive experiments and ablation study to demonstrate the effectiveness of our proposed method.

Ke Chen, Liangyan Li, Huan Liu et al.

Stereo Image Super-Resolution (stereoSR) has attracted significant attention in recent years due to the extensive deployment of dual cameras in mobile phones, autonomous vehicles and robots. In this work, we propose a new StereoSR method, named SwinFSR, based on an extension of SwinIR, originally designed for single image restoration, and the frequency domain knowledge obtained by the Fast Fourier Convolution (FFC). Specifically, to effectively gather global information, we modify the Residual Swin Transformer blocks (RSTBs) in SwinIR by explicitly incorporating the frequency domain knowledge using the FFC and employing the resulting residual Swin Fourier Transformer blocks (RSFTBs) for feature extraction. Besides, for the efficient and accurate fusion of stereo views, we propose a new cross-attention module referred to as RCAM, which achieves highly competitive performance while requiring less computational cost than the state-of-the-art cross-attention modules. Extensive experimental results and ablation studies demonstrate the effectiveness and efficiency of our proposed SwinFSR.

Yangyi Liu, Stefano Rini, Sadaf Salehkalaibar et al.

In federated learning (FL), the communication constraint between the remote learners and the Parameter Server (PS) is a crucial bottleneck. For this reason, model updates must be compressed so as to minimize the loss in accuracy resulting from the communication constraint. This paper proposes ``\emph{${\bf M}$-magnitude weighted $L_{\bf 2}$ distortion + $\bf 2$ degrees of freedom''} (M22) algorithm, a rate-distortion inspired approach to gradient compression for federated training of deep neural networks (DNNs). In particular, we propose a family of distortion measures between the original gradient and the reconstruction we referred to as ``$M$-magnitude weighted $L_2$'' distortion, and we assume that gradient updates follow an i.i.d. distribution -- generalized normal or Weibull, which have two degrees of freedom. In both the distortion measure and the gradient, there is one free parameter for each that can be fitted as a function of the iteration number. Given a choice of gradient distribution and distortion measure, we design the quantizer minimizing the expected distortion in gradient reconstruction. To measure the gradient compression performance under a communication constraint, we define the \emph{per-bit accuracy} as the optimal improvement in accuracy that one bit of communication brings to the centralized model over the training period. Using this performance measure, we systematically benchmark the choice of gradient distribution and distortion measure. We provide substantial insights on the role of these choices and argue that significant performance improvements can be attained using such a rate-distortion inspired compressor.

Jun Chen, Hong Chen, Xue Jiang et al.

Triplet learning, i.e. learning from triplet data, has attracted much attention in computer vision tasks with an extremely large number of categories, e.g., face recognition and person re-identification. Albeit with rapid progress in designing and applying triplet learning algorithms, there is a lacking study on the theoretical understanding of their generalization performance. To fill this gap, this paper investigates the generalization guarantees of triplet learning by leveraging the stability analysis. Specifically, we establish the first general high-probability generalization bound for the triplet learning algorithm satisfying the uniform stability, and then obtain the excess risk bounds of the order $O(n^{-\frac{1}{2}} \mathrm{log}n)$ for both stochastic gradient descent (SGD) and regularized risk minimization (RRM), where $2n$ is approximately equal to the number of training samples. Moreover, an optimistic generalization bound in expectation as fast as $O(n^{-1})$ is derived for RRM in a low noise case via the on-average stability analysis. Finally, our results are applied to triplet metric learning to characterize its theoretical underpinning.

Jiahuan Wang, Jun Chen, Hong Chen et al.

Recently, some mixture algorithms of pointwise and pairwise learning (PPL) have been formulated by employing the hybrid error metric of "pointwise loss + pairwise loss" and have shown empirical effectiveness on feature selection, ranking and recommendation tasks. However, to the best of our knowledge, the learning theory foundation of PPL has not been touched in the existing works. In this paper, we try to fill this theoretical gap by investigating the generalization properties of PPL. After extending the definitions of algorithmic stability to the PPL setting, we establish the high-probability generalization bounds for uniformly stable PPL algorithms. Moreover, explicit convergence rates of stochastic gradient descent (SGD) and regularized risk minimization (RRM) for PPL are stated by developing the stability analysis technique of pairwise learning. In addition, the refined generalization bounds of PPL are obtained by replacing uniform stability with on-average stability.

Xudong Gao, Xiao Guang Gao, Jia Rong et al.

Fuzzy General Grey Cognitive Map (FGGCM) and Fuzzy Grey Cognitive Map (FGCM) are extensions of Fuzzy Cognitive Map (FCM) in terms of uncertainty. FGGCM allows for the processing of general grey number with multiple intervals, enabling FCM to better address uncertain situations. Although the convergence of FCM and FGCM has been discussed in many literature, the convergence of FGGCM has not been thoroughly explored. This paper aims to fill this research gap. First, metrics for the general grey number space and its vector space is given and proved using the Minkowski inequality. By utilizing the characteristic that Cauchy sequences are convergent sequences, the completeness of these two space is demonstrated. On this premise, utilizing Banach fixed point theorem and Browder-Gohde-Kirk fixed point theorem, combined with Lagrange's mean value theorem and Cauchy's inequality, deduces the sufficient conditions for FGGCM to converge to a unique fixed point when using tanh and sigmoid functions as activation functions. The sufficient conditions for the kernels and greyness of FGGCM to converge to a unique fixed point are also provided separately. Finally, based on Web Experience and Civil engineering FCM, designed corresponding FGGCM with sigmoid and tanh as activation functions by modifying the weights to general grey numbers. By comparing with the convergence theorems of FCM and FGCM, the effectiveness of the theorems proposed in this paper was verified. It was also demonstrated that the convergence theorems of FCM are special cases of the theorems proposed in this paper. The study for convergence of FGGCM is of great significance for guiding the learning algorithm of FGGCM, which is needed for designing FGGCM with specific fixed points, lays a solid theoretical foundation for the application of FGGCM in fields such as control, prediction, and decision support systems.

Xintian Shen, Jiangning Zhang, Jun Chen et al.

Image harmonization aims to solve the visual inconsistency problem in composited images by adaptively adjusting the foreground pixels with the background as references. Existing methods employ local color transformation or region matching between foreground and background, which neglects powerful proximity prior and independently distinguishes fore-/back-ground as a whole part for harmonization. As a result, they still show a limited performance across varied foreground objects and scenes. To address this issue, we propose a novel Global-aware Kernel Network (GKNet) to harmonize local regions with comprehensive consideration of long-distance background references. Specifically, GKNet includes two parts, \ie, harmony kernel prediction and harmony kernel modulation branches. The former includes a Long-distance Reference Extractor (LRE) to obtain long-distance context and Kernel Prediction Blocks (KPB) to predict multi-level harmony kernels by fusing global information with local features. To achieve this goal, a novel Selective Correlation Fusion (SCF) module is proposed to better select relevant long-distance background references for local harmonization. The latter employs the predicted kernels to harmonize foreground regions with both local and global awareness. Abundant experiments demonstrate the superiority of our method for image harmonization over state-of-the-art methods, \eg, achieving 39.53dB PSNR that surpasses the best counterpart by +0.78dB $\uparrow$; decreasing fMSE/MSE by 11.5\%$\downarrow$/6.7\%$\downarrow$ compared with the SoTA method. Code will be available at \href{https://github.com/XintianShen/GKNet}{here}.

Ren Yang, Radu Timofte, Jing Liu et al.

This paper reviews the first NTIRE challenge on quality enhancement of compressed video, with a focus on the proposed methods and results. In this challenge, the new Large-scale Diverse Video (LDV) dataset is employed. The challenge has three tracks. Tracks 1 and 2 aim at enhancing the videos compressed by HEVC at a fixed QP, while Track 3 is designed for enhancing the videos compressed by x265 at a fixed bit-rate. Besides, the quality enhancement of Tracks 1 and 3 targets at improving the fidelity (PSNR), and Track 2 targets at enhancing the perceptual quality. The three tracks totally attract 482 registrations. In the test phase, 12 teams, 8 teams and 11 teams submitted the final results of Tracks 1, 2 and 3, respectively. The proposed methods and solutions gauge the state-of-the-art of video quality enhancement. The homepage of the challenge: https://github.com/RenYang-home/NTIRE21_VEnh

Jiangning Zhang, Xianfang Zeng, Chao Xu et al.

Audio-guided face reenactment aims to generate a photorealistic face that has matched facial expression with the input audio. However, current methods can only reenact a special person once the model is trained or need extra operations such as 3D rendering and image post-fusion on the premise of generating vivid faces. To solve the above challenge, we propose a novel \emph{R}eal-time \emph{A}udio-guided \emph{M}ulti-face reenactment approach named \emph{APB2FaceV2}, which can reenact different target faces among multiple persons with corresponding reference face and drive audio signal as inputs. Enabling the model to be trained end-to-end and have a faster speed, we design a novel module named Adaptive Convolution (AdaConv) to infuse audio information into the network, as well as adopt a lightweight network as our backbone so that the network can run in real time on CPU and GPU. Comparison experiments prove the superiority of our approach than existing state-of-the-art methods, and further experiments demonstrate that our method is efficient and flexible for practical applications https://github.com/zhangzjn/APB2FaceV2

Linhui Dai, Xiaohong Liu, Chengqi Li et al.

As the revolutionary improvement being made on the performance of smartphones over the last decade, mobile photography becomes one of the most common practices among the majority of smartphone users. However, due to the limited size of camera sensors on phone, the photographed image is still visually distinct to the one taken by the digital single-lens reflex (DSLR) camera. To narrow this performance gap, one is to redesign the camera image signal processor (ISP) to improve the image quality. Owing to the rapid rise of deep learning, recent works resort to the deep convolutional neural network (CNN) to develop a sophisticated data-driven ISP that directly maps the phone-captured image to the DSLR-captured one. In this paper, we introduce a novel network that utilizes the attention mechanism and wavelet transform, dubbed AWNet, to tackle this learnable image ISP problem. By adding the wavelet transform, our proposed method enables us to restore favorable image details from RAW information and achieve a larger receptive field while remaining high efficiency in terms of computational cost. The global context block is adopted in our method to learn the non-local color mapping for the generation of appealing RGB images. More importantly, this block alleviates the influence of image misalignment occurred on the provided dataset. Experimental results indicate the advances of our design in both qualitative and quantitative measurements. The code is available publically.

Zheng Liu, Botao Xiao, Muhammad Alrabeiah et al.

Haze and smog are among the most common environmental factors impacting image quality and, therefore, image analysis. This paper proposes an end-to-end generative method for image dehazing. It is based on designing a fully convolutional neural network to recognize haze structures in input images and restore clear, haze-free images. The proposed method is agnostic in the sense that it does not explore the atmosphere scattering model. Somewhat surprisingly, it achieves superior performance relative to all existing state-of-the-art methods for image dehazing even on SOTS outdoor images, which are synthesized using the atmosphere scattering model. Project detail and code can be found here: https://github.com/Seanforfun/GMAN_Net_Haze_Removal

Marcos V. Conde, Florin-Alexandru Vasluianu, Radu Timofte et al.

This paper reviews the NTIRE 2024 RAW Image Super-Resolution Challenge, highlighting the proposed solutions and results. New methods for RAW Super-Resolution could be essential in modern Image Signal Processing (ISP) pipelines, however, this problem is not as explored as in the RGB domain. Th goal of this challenge is to upscale RAW Bayer images by 2x, considering unknown degradations such as noise and blur. In the challenge, a total of 230 participants registered, and 45 submitted results during thee challenge period. The performance of the top-5 submissions is reviewed and provided here as a gauge for the current state-of-the-art in RAW Image Super-Resolution.

Sadaf Salehkalaibar, Jun Chen, Ashish Khisti et al.

This paper studies the rate-distortion-perception (RDP) tradeoff for a memoryless source model in the asymptotic limit of large block-lengths. The perception measure is based on a divergence between the distributions of the source and reconstruction sequences \emph{conditioned} on the encoder output, first proposed by Mentzer et al. We consider the case when there is no shared randomness between the encoder and the decoder and derive a single-letter characterization of the RDP function for the case of discrete memoryless sources. This is in contrast to the marginal-distribution metric case (introduced by Blau and Michaeli), whose RDP characterization remains open when there is no shared randomness. The achievability scheme is based on lossy source coding with a posterior reference map. For the case of continuous valued sources under the squared error distortion measure and the squared quadratic Wasserstein perception measure, we also derive a single-letter characterization and show that the decoder can be restricted to a noise-adding mechanism. Interestingly, the RDP function characterized for the case of zero perception loss coincides with that of the marginal metric, and further zero perception loss can be achieved with a 3-dB penalty in minimum distortion. Finally we specialize to the case of Gaussian sources, and derive the RDP function for Gaussian vector case and propose a reverse water-filling type solution. We also partially characterize the RDP function for a mixture of Gaussian vector sources.

Xiqiang Qu, Jun Chen, Lei Yu et al.

We establish a single-letter characterization of the fundamental distortion-rate-perception tradeoff with limited common randomness under the squared error distortion measure and the squared Wasserstein-2 perception measure. Moreover, it is shown that this single-letter characterization can be explicitly evaluated for the Gaussian source. Various notions of universal representation are also clarified.

Kai Han, Siqi Ma, Chengxuan Qian et al.

Accurate segmentation of tumors and adjacent normal tissues in medical images is essential for surgical planning and tumor staging. Although foundation models generally perform well in segmentation tasks, they often struggle to focus on foreground areas in complex, low-contrast backgrounds, where some malignant tumors closely resemble normal organs, complicating contextual differentiation. To address these challenges, we propose the Foreground-Aware Spectrum Segmentation (FASS) framework. First, we introduce a foreground-aware module to amplify the distinction between background and the entire volume space, allowing the model to concentrate more effectively on target areas. Next, a feature-level frequency enhancement module, based on wavelet transform, extracts discriminative high-frequency features to enhance boundary recognition and detail perception. Eventually, we introduce an edge constraint module to preserve geometric continuity in segmentation boundaries. Extensive experiments on multiple medical datasets demonstrate superior performance across all metrics, validating the effectiveness of our framework, particularly in robustness under complex conditions and fine structure recognition. Our framework significantly enhances segmentation of low-contrast images, paving the way for applications in more diverse and complex medical imaging scenarios.

Anwen Lu, Mingxin Liu, Yiping Jiao et al.

Primary liver malignancies are widely recognized as the most heterogeneous and prognostically diverse cancers of the digestive system. Among these, hepatocellular carcinoma (HCC) and intrahepatic cholangiocarcinoma (ICC) emerge as the two principal histological subtypes, demonstrating significantly greater complexity in tissue morphology and cellular architecture than other common tumors. The intricate representation of features in Whole Slide Images (WSIs) encompasses abundant crucial information for liver cancer histological subtyping, regarding hierarchical pyramid structure, tumor microenvironment (TME), and geometric representation. However, recent approaches have not adequately exploited these indispensable effective descriptors, resulting in a limited understanding of histological representation and suboptimal subtyping performance. To mitigate these limitations, ARGUS is proposed to advance histological subtyping in liver cancer by capturing the macro-meso-micro hierarchical information within the TME. Specifically, we first construct a micro-geometry feature to represent fine-grained cell-level pattern via a geometric structure across nuclei, thereby providing a more refined and precise perspective for delineating pathological images. Then, a Hierarchical Field-of-Views (FoVs) Alignment module is designed to model macro- and meso-level hierarchical interactions inherent in WSIs. Finally, the augmented micro-geometry and FoVs features are fused into a joint representation via present Geometry Prior Guided Fusion strategy for modeling holistic phenotype interactions. Extensive experiments on public and private cohorts demonstrate that our ARGUS achieves state-of-the-art (SOTA) performance in histological subtyping of liver cancer, which provide an effective diagnostic tool for primary liver malignancies in clinical practice.

Jingyang Xiang, Zhuangzhi Chen, Jianbiao Mei et al.

Soft filter pruning~(SFP) has emerged as an effective pruning technique for allowing pruned filters to update and the opportunity for them to regrow to the network. However, this pruning strategy applies training and pruning in an alternative manner, which inevitably causes inconsistent representations between the reconstructed network~(R-NN) at the training and the pruned network~(P-NN) at the inference, resulting in performance degradation. In this paper, we propose to mitigate this gap by learning consistent representation for soft filter pruning, dubbed as CR-SFP. Specifically, for each training step, CR-SFP optimizes the R-NN and P-NN simultaneously with different distorted versions of the same training data, while forcing them to be consistent by minimizing their posterior distribution via the bidirectional KL-divergence loss. Meanwhile, the R-NN and P-NN share backbone parameters thus only additional classifier parameters are introduced. After training, we can export the P-NN for inference. CR-SFP is a simple yet effective training framework to improve the accuracy of P-NN without introducing any additional inference cost. It can also be combined with a variety of pruning criteria and loss functions. Extensive experiments demonstrate our CR-SFP achieves consistent improvements across various CNN architectures. Notably, on ImageNet, our CR-SFP reduces more than 41.8\% FLOPs on ResNet18 with 69.2\% top-1 accuracy, improving SFP by 2.1\% under the same training settings. The code will be publicly available on GitHub.

Sadaf Salehkalaibar, Buu Phan, Jun Chen et al.

We study causal, low-latency, sequential video compression when the output is subjected to both a mean squared-error (MSE) distortion loss as well as a perception loss to target realism. Motivated by prior approaches, we consider two different perception loss functions (PLFs). The first, PLF-JD, considers the joint distribution (JD) of all the video frames up to the current one, while the second metric, PLF-FMD, considers the framewise marginal distributions (FMD) between the source and reconstruction. Using information theoretic analysis and deep-learning based experiments, we demonstrate that the choice of PLF can have a significant effect on the reconstruction, especially at low-bit rates. In particular, while the reconstruction based on PLF-JD can better preserve the temporal correlation across frames, it also imposes a significant penalty in distortion compared to PLF-FMD and further makes it more difficult to recover from errors made in the earlier output frames. Although the choice of PLF decisively affects reconstruction quality, we also demonstrate that it may not be essential to commit to a particular PLF during encoding and the choice of PLF can be delegated to the decoder. In particular, encoded representations generated by training a system to minimize the MSE (without requiring either PLF) can be {\em near universal} and can generate close to optimal reconstructions for either choice of PLF at the decoder. We validate our results using (one-shot) information-theoretic analysis, detailed study of the rate-distortion-perception tradeoff of the Gauss-Markov source model as well as deep-learning based experiments on moving MNIST and KTH datasets.

Jun Chen, Weng-Keen Wong, Bechir Hamdaoui et al.

Recent deep neural network-based device classification studies show that complex-valued neural networks (CVNNs) yield higher classification accuracy than real-valued neural networks (RVNNs). Although this improvement is (intuitively) attributed to the complex nature of the input RF data (i.e., IQ symbols), no prior work has taken a closer look into analyzing such a trend in the context of wireless device identification. Our study provides a deeper understanding of this trend using real LoRa and WiFi RF datasets. We perform a deep dive into understanding the impact of (i) the input representation/type and (ii) the architectural layer of the neural network. For the input representation, we considered the IQ as well as the polar coordinates both partially and fully. For the architectural layer, we considered a series of ablation experiments that eliminate parts of the CVNN components. Our results show that CVNNs consistently outperform RVNNs counterpart in the various scenarios mentioned above, indicating that CVNNs are able to make better use of the joint information provided via the in-phase (I) and quadrature (Q) components of the signal.

Jun Chen, Yuang Liu, Xiangrui Zhao et al.

In a Riemannian manifold, the Ricci flow is a partial differential equation for evolving the metric to become more regular. We hope that topological structures from such metrics may be used to assist in the tasks of machine learning. However, this part of the work is still missing. In this paper, we propose Ricci flow assisted Eucl2Hyp2Eucl neural networks that bridge this gap between the Ricci flow and deep neural networks by mapping neural manifolds from the Euclidean space to the dynamically stable Poincaré ball and then back to the Euclidean space. As a result, we prove that, if initial metrics have an $L^2$-norm perturbation which deviates from the Hyperbolic metric on the Poincaré ball, the scaled Ricci-DeTurck flow of such metrics smoothly and exponentially converges to the Hyperbolic metric. Specifically, the role of the Ricci flow is to serve as naturally evolving to the stable Poincaré ball. For such dynamically stable neural manifolds under the Ricci flow, the convergence of neural networks embedded with such manifolds is not susceptible to perturbations. And we show that Ricci flow assisted Eucl2Hyp2Eucl neural networks outperform with their all Euclidean counterparts on image classification tasks.

Zhihao Shi, Xiangyu Xu, Xiaohong Liu et al.

Existing methods for video interpolation heavily rely on deep convolution neural networks, and thus suffer from their intrinsic limitations, such as content-agnostic kernel weights and restricted receptive field. To address these issues, we propose a Transformer-based video interpolation framework that allows content-aware aggregation weights and considers long-range dependencies with the self-attention operations. To avoid the high computational cost of global self-attention, we introduce the concept of local attention into video interpolation and extend it to the spatial-temporal domain. Furthermore, we propose a space-time separation strategy to save memory usage, which also improves performance. In addition, we develop a multi-scale frame synthesis scheme to fully realize the potential of Transformers. Extensive experiments demonstrate the proposed model performs favorably against the state-of-the-art methods both quantitatively and qualitatively on a variety of benchmark datasets.

George Zhang, Jingjing Qian, Jun Chen et al.

In the context of lossy compression, Blau & Michaeli (2019) adopt a mathematical notion of perceptual quality and define the information rate-distortion-perception function, generalizing the classical rate-distortion tradeoff. We consider the notion of universal representations in which one may fix an encoder and vary the decoder to achieve any point within a collection of distortion and perception constraints. We prove that the corresponding information-theoretic universal rate-distortion-perception function is operationally achievable in an approximate sense. Under MSE distortion, we show that the entire distortion-perception tradeoff of a Gaussian source can be achieved by a single encoder of the same rate asymptotically. We then characterize the achievable distortion-perception region for a fixed representation in the case of arbitrary distributions, identify conditions under which the aforementioned results continue to hold approximately, and study the case when the rate is not fixed in advance. This motivates the study of practical constructions that are approximately universal across the RDP tradeoff, thereby alleviating the need to design a new encoder for each objective. We provide experimental results on MNIST and SVHN suggesting that on image compression tasks, the operational tradeoffs achieved by machine learning models with a fixed encoder suffer only a small penalty when compared to their variable encoder counterparts.

Yankun Yu, Huan Liu, Minghan Fu et al.

Recently, there has been rapid and significant progress on image dehazing. Many deep learning based methods have shown their superb performance in handling homogeneous dehazing problems. However, we observe that even if a carefully designed convolutional neural network (CNN) can perform well on large-scaled dehazing benchmarks, the network usually fails on the non-homogeneous dehazing datasets introduced by NTIRE challenges. The reasons are mainly in two folds. Firstly, due to its non-homogeneous nature, the non-uniformly distributed haze is harder to be removed than the homogeneous haze. Secondly, the research challenge only provides limited data (there are only 25 training pairs in NH-Haze 2021 dataset). Thus, learning the mapping from the domain of hazy images to that of clear ones based on very limited data is extremely hard. To this end, we propose a simple but effective approach for non-homogeneous dehazing via ensemble learning. To be specific, we introduce a two-branch neural network to separately deal with the aforementioned problems and then map their distinct features by a learnable fusion tail. We show extensive experimental results to illustrate the effectiveness of our proposed method.

Xiaohong Liu, Yongrui Ma, Zhihao Shi et al.

We develop a new physical model for the rain effect and show that the well-known atmosphere scattering model (ASM) for the haze effect naturally emerges as its homogeneous continuous limit. Via depth-aware fusion of multi-layer rain streaks according to the camera imaging mechanism, the new model can better capture the sophisticated non-deterministic degradation patterns commonly seen in real rainy images. We also propose a Densely Scale-Connected Attentive Network (DSCAN) that is suitable for both deraining and dehazing tasks. Our design alleviates the bottleneck issue existent in conventional multi-scale networks and enables more effective information exchange and aggregation. Extensive experimental results demonstrate that the proposed DSCAN is able to deliver superior derained/dehazed results on both synthetic and real images as compared to the state-of-the-art. Moreover, it is shown that for our DSCAN, the synthetic dataset built using the new physical model yields better generalization performance on real images in comparison with the existing datasets based on over-simplified models.

Xiaohong Liu, Zhihao Shi, Zijun Wu et al.

We propose an enhanced multi-scale network, dubbed GridDehazeNet+, for single image dehazing. The proposed dehazing method does not rely on the Atmosphere Scattering Model (ASM), and an explanation as to why it is not necessarily performing the dimension reduction offered by this model is provided. GridDehazeNet+ consists of three modules: pre-processing, backbone, and post-processing. The trainable pre-processing module can generate learned inputs with better diversity and more pertinent features as compared to those derived inputs produced by hand-selected pre-processing methods. The backbone module implements multi-scale estimation with two major enhancements: 1) a novel grid structure that effectively alleviates the bottleneck issue via dense connections across different scales; 2) a spatial-channel attention block that can facilitate adaptive fusion by consolidating dehazing-relevant features. The post-processing module helps to reduce the artifacts in the final output. Due to domain shift, the model trained on synthetic data may not generalize well on real data. To address this issue, we shape the distribution of synthetic data to match that of real data, and use the resulting translated data to finetune our network. We also propose a novel intra-task knowledge transfer mechanism that can memorize and take advantage of synthetic domain knowledge to assist the learning process on the translated data. Experimental results demonstrate that the proposed method outperforms the state-of-the-art on several synthetic dehazing datasets, and achieves the superior performance on real-world hazy images after finetuning.

Xiaohong Liu, Yaojie Liu, Jun Chen et al.

To defend against manipulation of image content, such as splicing, copy-move, and removal, we develop a Progressive Spatio-Channel Correlation Network (PSCC-Net) to detect and localize image manipulations. PSCC-Net processes the image in a two-path procedure: a top-down path that extracts local and global features and a bottom-up path that detects whether the input image is manipulated, and estimates its manipulation masks at multiple scales, where each mask is conditioned on the previous one. Different from the conventional encoder-decoder and no-pooling structures, PSCC-Net leverages features at different scales with dense cross-connections to produce manipulation masks in a coarse-to-fine fashion. Moreover, a Spatio-Channel Correlation Module (SCCM) captures both spatial and channel-wise correlations in the bottom-up path, which endows features with holistic cues, enabling the network to cope with a wide range of manipulation attacks. Thanks to the light-weight backbone and progressive mechanism, PSCC-Net can process 1,080P images at 50+ FPS. Extensive experiments demonstrate the superiority of PSCC-Net over the state-of-the-art methods on both detection and localization.

Zhihao Shi, Xiaohong Liu, Chengqi Li et al.

Recent years have seen considerable research activities devoted to video enhancement that simultaneously increases temporal frame rate and spatial resolution. However, the existing methods either fail to explore the intrinsic relationship between temporal and spatial information or lack flexibility in the choice of final temporal/spatial resolution. In this work, we propose an unconstrained space-time video super-resolution network, which can effectively exploit space-time correlation to boost performance. Moreover, it has complete freedom in adjusting the temporal frame rate and spatial resolution through the use of the optical flow technique and a generalized pixelshuffle operation. Our extensive experiments demonstrate that the proposed method not only outperforms the state-of-the-art, but also requires far fewer parameters and less running time.

Nima Shoghi Ghalehshahi, Ramyad Hadidi, Lee Jaewon et al.

Intelligent transportation systems (ITS) are expected to effectively create a stand-alone network for secure communication among autonomous agents. In such a dynamic and fast-changing network with high-speed agents, verifying the authenticity and integrity of messages while taking preventive action (e.g., applying brakes) within tens of milliseconds is one of the main challenges. In such a brief moment after receiving a message, the agent not only must verify the integrity and authenticity of the received message but also needs to perform extra computations to localize the sender of the message for taking appropriate action (e.g., an immediate stop warning from a vehicle in front vs. rear). In this paper, we present an inherently location-aware and lightweight authentication protocol by exploiting in situ visual localization (i.e., SLAM). In this protocol, each agent displays its public key using visual authentication beacons (e.g., QR codes). Thus, receiving agents not only can verify and authenticate the messages but also can easily localize the sender by keeping a shortlist of observed visual beacons within their visual localization system with no additional computation cost. Compared to prior work, our location-aware protocol is scalable, does not depend on any infrastructure, removes the high cost of post-message-delivery localization, and provides trustworthiness guarantees for information that are beyond the reach of each agent sensors.

Andrey Ignatov, Radu Timofte, Zhilu Zhang et al.

This paper reviews the second AIM learned ISP challenge and provides the description of the proposed solutions and results. The participating teams were solving a real-world RAW-to-RGB mapping problem, where to goal was to map the original low-quality RAW images captured by the Huawei P20 device to the same photos obtained with the Canon 5D DSLR camera. The considered task embraced a number of complex computer vision subtasks, such as image demosaicing, denoising, white balancing, color and contrast correction, demoireing, etc. The target metric used in this challenge combined fidelity scores (PSNR and SSIM) with solutions' perceptual results measured in a user study. The proposed solutions significantly improved the baseline results, defining the state-of-the-art for practical image signal processing pipeline modeling.

Zhihao Shi, Xiaohong Liu, Kangdi Shi et al.

Video frame interpolation aims at synthesizing intermediate frames from nearby source frames while maintaining spatial and temporal consistencies. The existing deep-learning-based video frame interpolation methods can be roughly divided into two categories: flow-based methods and kernel-based methods. The performance of flow-based methods is often jeopardized by the inaccuracy of flow map estimation due to oversimplified motion models, while that of kernel-based methods tends to be constrained by the rigidity of kernel shape. To address these performance-limiting issues, a novel mechanism named generalized deformable convolution is proposed, which can effectively learn motion information in a data-driven manner and freely select sampling points in space-time. We further develop a new video frame interpolation method based on this mechanism. Our extensive experiments demonstrate that the new method performs favorably against the state-of-the-art, especially when dealing with complex motions.

Shanxin Yuan, Radu Timofte, Ales Leonardis et al.

This paper reviews the Challenge on Image Demoireing that was part of the New Trends in Image Restoration and Enhancement (NTIRE) workshop, held in conjunction with CVPR 2020. Demoireing is a difficult task of removing moire patterns from an image to reveal an underlying clean image. The challenge was divided into two tracks. Track 1 targeted the single image demoireing problem, which seeks to remove moire patterns from a single image. Track 2 focused on the burst demoireing problem, where a set of degraded moire images of the same scene were provided as input, with the goal of producing a single demoired image as output. The methods were ranked in terms of their fidelity, measured using the peak signal-to-noise ratio (PSNR) between the ground truth clean images and the restored images produced by the participants' methods. The tracks had 142 and 99 registered participants, respectively, with a total of 14 and 6 submissions in the final testing stage. The entries span the current state-of-the-art in image and burst image demoireing problems.

Xiaohong Liu, Lingshi Kong, Yang Zhou et al.

Video super-resolution aims at generating a high-resolution video from its low-resolution counterpart. With the rapid rise of deep learning, many recently proposed video super-resolution methods use convolutional neural networks in conjunction with explicit motion compensation to capitalize on statistical dependencies within and across low-resolution frames. Two common issues of such methods are noteworthy. Firstly, the quality of the final reconstructed HR video is often very sensitive to the accuracy of motion estimation. Secondly, the warp grid needed for motion compensation, which is specified by the two flow maps delineating pixel displacements in horizontal and vertical directions, tends to introduce additional errors and jeopardize the temporal consistency across video frames. To address these issues, we propose a novel dynamic local filter network to perform implicit motion estimation and compensation by employing, via locally connected layers, sample-specific and position-specific dynamic local filters that are tailored to the target pixels. We also propose a global refinement network based on ResBlock and autoencoder structures to exploit non-local correlations and enhance the spatial consistency of super-resolved frames. The experimental results demonstrate that the proposed method outperforms the state-of-the-art, and validate its strength in terms of local transformation handling, temporal consistency as well as edge sharpness.

Xiaohong Liu, Yongrui Ma, Zhihao Shi et al.

We propose an end-to-end trainable Convolutional Neural Network (CNN), named GridDehazeNet, for single image dehazing. The GridDehazeNet consists of three modules: pre-processing, backbone, and post-processing. The trainable pre-processing module can generate learned inputs with better diversity and more pertinent features as compared to those derived inputs produced by hand-selected pre-processing methods. The backbone module implements a novel attention-based multi-scale estimation on a grid network, which can effectively alleviate the bottleneck issue often encountered in the conventional multi-scale approach. The post-processing module helps to reduce the artifacts in the final output. Experimental results indicate that the GridDehazeNet outperforms the state-of-the-arts on both synthetic and real-world images. The proposed hazing method does not rely on the atmosphere scattering model, and we provide an explanation as to why it is not necessarily beneficial to take advantage of the dimension reduction offered by the atmosphere scattering model for image dehazing, even if only the dehazing results on synthetic images are concerned.

Chao Tian, Hua Sun, Jun Chen

We propose a new capacity-achieving code for the private information retrieval (PIR) problem, and show that it has the minimum message size (being one less than the number of servers) and the minimum upload cost (being roughly linear in the number of messages) among a general class of capacity-achieving codes, and in particular, among all capacity-achieving linear codes. Different from existing code constructions, the proposed code is asymmetric, and this asymmetry appears to be the key factor leading to the optimal message size and the optimal upload cost. The converse results on the message size and the upload cost are obtained by a strategic analysis of the information theoretic proof of the PIR capacity, from which a set of critical properties of any capacity-achieving code in the code class of interest is extracted. The symmetry structure of the PIR problem is then analyzed, which allows us to construct symmetric codes from asymmetric ones, yielding a meaningful bridge between the proposed code and existing ones in the literature.

Yang Xiao, Jun Chen, Yancheng Wang et al.

Dynamic imaging is a recently proposed action description paradigm for simultaneously capturing motion and temporal evolution information, particularly in the context of deep convolutional neural networks (CNNs). Compared with optical flow for motion characterization, dynamic imaging exhibits superior efficiency and compactness. Inspired by the success of dynamic imaging in RGB video, this study extends it to the depth domain. To better exploit three-dimensional (3D) characteristics, multi-view dynamic images are proposed. In particular, the raw depth video is densely projected with respect to different virtual imaging viewpoints by rotating the virtual camera within the 3D space. Subsequently, dynamic images are extracted from the obtained multi-view depth videos and multi-view dynamic images are thus constructed from these images. Accordingly, more view-tolerant visual cues can be involved. A novel CNN model is then proposed to perform feature learning on multi-view dynamic images. Particularly, the dynamic images from different views share the same convolutional layers but correspond to different fully connected layers. This is aimed at enhancing the tuning effectiveness on shallow convolutional layers by alleviating the gradient vanishing problem. Moreover, as the spatial occurrence variation of the actions may impair the CNN, an action proposal approach is also put forth. In experiments, the proposed approach can achieve state-of-the-art performance on three challenging datasets.

Jun Chen, Guang Yang, Zhifan Gao et al.

Late Gadolinium Enhanced Cardiac MRI (LGE-CMRI) for detecting atrial scars in atrial fibrillation (AF) patients has recently emerged as a promising technique to stratify patients, guide ablation therapy and predict treatment success. Visualisation and quantification of scar tissues require a segmentation of both the left atrium (LA) and the high intensity scar regions from LGE-CMRI images. These two segmentation tasks are challenging due to the cancelling of healthy tissue signal, low signal-to-noise ratio and often limited image quality in these patients. Most approaches require manual supervision and/or a second bright-blood MRI acquisition for anatomical segmentation. Segmenting both the LA anatomy and the scar tissues automatically from a single LGE-CMRI acquisition is highly in demand. In this study, we proposed a novel fully automated multiview two-task (MVTT) recursive attention model working directly on LGE-CMRI images that combines a sequential learning and a dilated residual learning to segment the LA (including attached pulmonary veins) and delineate the atrial scars simultaneously via an innovative attention model. Compared to other state-of-the-art methods, the proposed MVTT achieves compelling improvement, enabling to generate a patient-specific anatomical and atrial scar assessment model.

Jun Chen, Keisuke Nonaka, Ryosuke Watanabe et al.

This paper proposes a new parallel approach to solve connected components on a 2D binary image implemented with CUDA. We employ the following strategies to accelerate neighborhood exploration after dividing an input image into independent blocks. In the local labeling stage, a coarse-labeling algorithm, including row-column connection and label-equivalence list unification, is applied first to sort out the mess of an initialized local label map; a refinement algorithm is then introduced to merge separated sub-regions from a single component. In the block merge stage, we scan the pixels located on the boundary of each block instead of solving the connectivity of all the pixels. With the proposed method, the length of label-equivalence lists is compressed, and the number of memory accesses is reduced. Thus, the efficiency of connected components labeling is improved. Experimental results show that our method outperforms the other approaches between $29\%$ and $80\%$ on average.

Jun Chen, Qiang Yao, Houari Sabirin et al.

In this paper, we report an optimized union-find (UF) algorithm that can label the connected components on a 2D image efficiently by employing the GPU architecture. The proposed method contains three phases: UF-based local merge, boundary analysis, and link. The coarse labeling in local merge reduces the number atomic operations, while the boundary analysis only manages the pixels on the boundary of each block. Evaluation results showed that the proposed algorithm speed up the average running time by more than 1.3X.