High-resolution digital images are usually downscaled to fit various display screens or save the cost of storage and bandwidth, meanwhile the post-upscaling is adpoted to recover the original resolutions or the details in the zoom-in images. However, typical image downscaling is a non-injective mapping due to the loss of high-frequency information, which leads to the ill-posed problem of the inverse upscaling procedure and poses great challenges for recovering details from the downscaled low-resolution images. Simply upscaling with image super-resolution methods results in unsatisfactory recovering performance. In this work, we propose to solve this problem by modeling the downscaling and upscaling processes from a new perspective, i.e. an invertible bijective transformation, which can largely mitigate the ill-posed nature of image upscaling. We develop an Invertible Rescaling Net (IRN) with deliberately designed framework and objectives to produce visually-pleasing low-resolution images and meanwhile capture the distribution of the lost information using a latent variable following a specified distribution in the downscaling process. In this way, upscaling is made tractable by inversely passing a randomly-drawn latent variable with the low-resolution image through the network. Experimental results demonstrate the significant improvement of our model over existing methods in terms of both quantitative and qualitative evaluations of image upscaling reconstruction from downscaled images.
Peking University（北京大学）；Microsoft Research Asia（微软亚洲研究院）；Peng Cheng Laboratory（鹏城实验室）
Image rescaling is a commonly used bidirectional operation, which first downscales high-resolution images to fit various display screens or to be storage- and bandwidth-friendly, and afterward upscales the corresponding low-resolution images to recover the original resolution or the details in the zoom-in images. However, the non-injective downscaling mapping discards high-frequency contents, leading to the ill-posed problem for the inverse restoration task. This can be abstracted as a general image degradation-restoration problem with information loss. In this work, we propose a novel inverti…
Great successes have been achieved using deep learning techniques for image super-resolution (SR) with fixed scales. To increase its real world applicability, numerous models have also been proposed to restore SR images with arbitrary scale factors, including asymmetric ones where images are resized to different scales along horizontal and vertical directions. Though most models are only optimized for the unidirectional upscaling task while assuming a predefined downscaling kernel for low-resolution (LR) inputs, recent models based on Invertible Neural Networks (INN) are able to increase upsca…
Bytedance Inc., Shenzhen, China（字节跳动有限公司（深圳））；Bytedance Inc., San Diego, CA, 92122 USA（字节跳动有限公司（圣迭戈））
High-resolution (HR) images are usually downscaled to low-resolution (LR) ones for better display and afterward upscaled back to the original size to recover details. Recent work in image rescaling formulates downscaling and upscaling as a unified task and learns a bijective mapping between HR and LR via invertible networks. However, in real-world applications (e.g., social media), most images are compressed for transmission. Lossy compression will lead to irreversible information loss on LR images, hence damaging the inverse upscaling procedure and degrading the reconstruction accuracy. In th…
University of Southern California（南加州大学）；Baidu Research (USA)（百度研究院（美国））
Normalizing flow models have been used successfully for generative image super-resolution (SR) by approximating complex distribution of natural images to simple tractable distribution in latent space through Invertible Neural Networks (INN). These models can generate multiple realistic SR images from one low-resolution (LR) input using randomly sampled points in the latent space, simulating the ill-posed nature of image upscaling where multiple high-resolution (HR) images correspond to the same LR. Lately, the invertible process in INN has also been used successfully by bidirectional image res…
Normalizing flow models using invertible neural networks (INN) have been widely investigated for successful generative image super-resolution (SR) by learning the transformation between the normal distribution of latent variable $z$ and the conditional distribution of high-resolution (HR) images gave a low-resolution (LR) input. Recently, image rescaling models like IRN utilize the bidirectional nature of INN to push the performance limit of image upscaling by optimizing the downscaling and upscaling steps jointly. While the random sampling of latent variable $z$ is useful in generating divers…
Computer Vision Lab, ETH Zurich（苏黎世联邦理工学院计算机视觉实验室）；KU Leuven（鲁汶大学）
Normalizing flows have recently demonstrated promising results for low-level vision tasks. For image super-resolution (SR), it learns to predict diverse photo-realistic high-resolution (HR) images from the low-resolution (LR) image rather than learning a deterministic mapping. For image rescaling, it achieves high accuracy by jointly modelling the downscaling and upscaling processes. While existing approaches employ specialized techniques for these two tasks, we set out to unify them in a single formulation. In this paper, we propose the hierarchical conditional flow (HCFlow) as a unified fram…
Deep learning based single image super-resolution models have been widely studied and superb results are achieved in upscaling low-resolution images with fixed scale factor and downscaling degradation kernel. To improve real world applicability of such models, there are growing interests to develop models optimized for arbitrary upscaling factors. Our proposed method is the first to treat arbitrary rescaling, both upscaling and downscaling, as one unified process. Using joint optimization of both directions, the proposed model is able to learn upscaling and downscaling simultaneously and achie…
Conventional social media platforms usually downscale the HR images to restrict their resolution to a specific size for saving transmission/storage cost, which leads to the super-resolution (SR) being highly ill-posed. Recent invertible image downscaling methods jointly model the downscaling/upscaling problems and achieve significant improvements. However, they only consider fixed integer scale factors that cannot downscale HR images with various resolutions to meet the resolution restriction of social media platforms. In this paper, we propose a scale-Arbitrary Invertible image Downscaling Ne…
Image denoising is a typical ill-posed problem due to complex degradation. Leading methods based on normalizing flows have tried to solve this problem with an invertible transformation instead of a deterministic mapping. However, the implicit bijective mapping is not explored well. Inspired by a latent observation that noise tends to appear in the high-frequency part of the image, we propose a fully invertible denoising method that injects the idea of disentangled learning into a general invertible neural network to split noise from the high-frequency part. More specifically, we decompose the …
Although deep learning based image compression methods have achieved promising progress these days, the performance of these methods still cannot match the latest compression standard Versatile Video Coding (VVC). Most of the recent developments focus on designing a more accurate and flexible entropy model that can better parameterize the distributions of the latent features. However, few efforts are devoted to structuring a better transformation between the image space and the latent feature space. In this paper, instead of employing previous autoencoder style networks to build this transform…
In this paper, we propose an invertible neural network for image denoising (DnINN) inspired by the transform-based denoising framework. The proposed DnINN consists of an invertible neural network called LINN whose architecture is inspired by the lifting scheme in wavelet theory and a sparsity-driven denoising network which is used to remove noise from the transform coefficients. The denoising operation is performed with a single soft-thresholding operation or with a learned iterative shrinkage thresholding network. The forward pass of LINN produces an over-complete representation which is more…
Existing unpaired low-light image enhancement approaches prefer to employ the two-way GAN framework, in which two CNN generators are deployed for enhancement and degradation separately. However, such data-driven models ignore the inherent characteristics of transformation between the low and normal light images, leading to unstable training and artifacts. Here, we propose to leverage the invertible network to enhance low-light image in forward process and degrade the normal-light one inversely with unpaired learning. The generated and real images are then fed into discriminators for adversaria…
Huazhong University of Science and Technology（华中科技大学）
Learning-based methods have effectively promoted the community of image compression. Meanwhile, variational autoencoder (VAE) based variable-rate approaches have recently gained much attention to avoid the usage of a set of different networks for various compression rates. Despite the remarkable performance that has been achieved, these approaches would be readily corrupted once multiple compression/decompression operations are executed, resulting in the fact that image quality would be tremendously dropped and strong artifacts would appear. Thus, we try to tackle the issue of high-fidelity fi…
Australian National University（澳大利亚国立大学）；Data61-CSIRO；OPPO US Research（OPPO美国研究院）；GSAI POSTECH（浦项科技大学GSAI）
Invertible networks have various benefits for image denoising since they are lightweight, information-lossless, and memory-saving during back-propagation. However, applying invertible models to remove noise is challenging because the input is noisy, and the reversed output is clean, following two different distributions. We propose an invertible denoising network, InvDN, to address this challenge. InvDN transforms the noisy input into a low-resolution clean image and a latent representation containing noise. To discard noise and restore the clean image, InvDN replaces the noisy latent represen…
Deep networks have achieved great success in image rescaling (IR) task that seeks to learn the optimal downscaled representations, i.e., low-resolution (LR) images, to reconstruct the original high-resolution (HR) images. Compared with super-resolution methods that consider a fixed downscaling scheme, e.g., bicubic, IR often achieves significantly better reconstruction performance thanks to the learned downscaled representations. This highlights the importance of a good downscaled representation in image reconstruction tasks. Existing IR methods mainly learn the downscaled representation by jo…
TVRN: Invertible Neural Networks for Compression-Aware Temporal Video Rescaling
MOE Key Laboratory of Brain-Inspired Intelligent Perception and Cognition（教育部脑科学与智能感知认知重点实验室）；University of Science and Technology of China（中国科学技术大学）；Information Science and Technology Institution（信息科学与技术研究所）；MCC Lab（MCC实验室）
Multi-Scale Invertible Neural Network for Wide-Range Variable-Rate Learned Image Compression
National Engineering Laboratory for Brain-Inspired Intelligence Technology and Application, University of Science and Technology of China（中国科学技术大学类脑智能技术及应用国家工程实验室）；University of Missouri（密苏里大学）
Microsoft Research Asia（微软亚洲研究院）；University of Toronto（多伦多大学）；Peking University（北京大学）
Lossy image compression is one of the most commonly used operators for digital images. Most recently proposed deep-learning-based image compression methods leverage the auto-encoder structure, and reach a series of promising results in this field. The images are encoded into low dimensional latent features first, and entropy coded subsequently by exploiting the statistical redundancy. However, the information lost during encoding is unfortunately inevitable, which poses a significant challenge to the decoder to reconstruct the original images. In this work, we propose a novel invertible framew…
Australian National University（澳大利亚国立大学）；Data61-CSIRO（澳大利亚联邦科学与工业研究组织数据61中心）
Ever since the advent of AlexNet, designing novel deep neural architectures for different tasks has consistently been a productive research direction. Despite the exceptional performance of various architectures in practice, we study a theoretical question: what is the condition for deep neural architectures to preserve all the information of the input data? Identifying the information lossless condition for deep neural architectures is important, because tasks such as image restoration require keep the detailed information of the input data as much as possible. Using the definition of mutual …
Image registration is a fundamental step in medical image analysis. Ideally, the transformation that registers one image to another should be a diffeomorphism that is both invertible and smooth. Traditional methods like geodesic shooting approach the problem via differential geometry, with theoretical guarantees that the resulting transformation will be smooth and invertible. Most previous research using unsupervised deep neural networks for registration have used a local smoothness constraint (typically, a spatial variation loss) to address the smoothness issue. These networks usually produce…
Nanyang Technological University（南洋理工大学）；City University of Hong Kong（香港城市大学）
To enhance low-light images to normally-exposed ones is highly ill-posed, namely that the mapping relationship between them is one-to-many. Previous works based on the pixel-wise reconstruction losses and deterministic processes fail to capture the complex conditional distribution of normally exposed images, which results in improper brightness, residual noise, and artifacts. In this paper, we investigate to model this one-to-many relationship via a proposed normalizing flow model. An invertible network that takes the low-light images/features as the condition and learns to map the distributio…
Joint Degradation-Aware Arbitrary-Scale Super-Resolution for Variable-Rate Extreme Image Compression
School of Information Science and Electronic Engineering, Shanghai Jiao Tong University（上海交通大学信息科学与电子工程学院）；Department of Electronic Engineer and Information Science, University of Science and Technology of China（中国科学技术大学电子工程与信息科学系）；MoE Key Lab of Artificial Intelligence, AI Institute, Shanghai Jiao Tong University（上海交通大学人工智能教育部重点实验室）
Crowd counting is a challenging problem especially in the presence of huge crowd diversity across images and complex cluttered crowd-like background regions, where most previous approaches do not generalize well and consequently produce either huge crowd underestimation or overestimation. To address these challenges, we propose a new image patch rescaling module (PRM) and three independent PRM employed crowd counting methods. The proposed frameworks use the PRM module to rescale the image regions (patches) that require special treatment, whereas the classification process helps in recognizing …
As an effective way to integrate the information contained in multiple medical images under different modalities, medical image synthesis and fusion have emerged in various clinical applications such as disease diagnosis and treatment planning. In this paper, an invertible and variable augmented network (iVAN) is proposed for medical image synthesis and fusion. In iVAN, the channel number of the network input and output is the same through variable augmentation technology, and data relevance is enhanced, which is conducive to the generation of characterization information. Meanwhile, the inver…
U-Nets have been established as a standard architecture for image-to-image learning problems such as segmentation and inverse problems in imaging. For large-scale data, as it for example appears in 3D medical imaging, the U-Net however has prohibitive memory requirements. Here, we present a new fully-invertible U-Net-based architecture called the iUNet, which employs novel learnable and invertible up- and downsampling operations, thereby making the use of memory-efficient backpropagation possible. This allows us to train deeper and larger networks in practice, under the same GPU memory restric…
Image downscaling and upscaling are two basic rescaling operations. Once the image is downscaled, it is difficult to be reconstructed via upscaling due to the loss of information. To make these two processes more compatible and improve the reconstruction performance, some efforts model them as a joint encoding-decoding task, with the constraint that the downscaled (i.e. encoded) low-resolution (LR) image must preserve the original visual appearance. To implement this constraint, most methods guide the downscaling module by supervising it with the bicubically downscaled LR version of the origin…
Tackling Ill-posedness of Reversible Image Conversion with Well-posed Invertible Network
School of Artificial Intelligence, Beijing Normal University（人工智能学院，北京师范大学）；Engineering Research Center of Intelligent Technology and Educational Application, Ministry of Education（智能技术及教育应用工程研究中心，教育部）
Indian Institute of Technology Kharagpur（印度理工学院卡拉格布尔分校）
The resurgence of deep neural networks has created an alternative pathway for low-dose computed tomography denoising by learning a nonlinear transformation function between low-dose CT (LDCT) and normal-dose CT (NDCT) image pairs. However, those paired LDCT and NDCT images are rarely available in the clinical environment, making deep neural network deployment infeasible. This study proposes a novel method for self-supervised low-dose CT denoising to alleviate the requirement of paired LDCT and NDCT images. Specifically, we have trained an invertible neural network to minimize the pixel-based m…
The Hong Kong University of Science and Technology（香港科技大学）
Unprocessed RAW data is a highly valuable image format for image editing and computer vision. However, since the file size of RAW data is huge, most users can only get access to processed and compressed sRGB images. To bridge this gap, we design an Invertible Image Signal Processing (InvISP) pipeline, which not only enables rendering visually appealing sRGB images but also allows recovering nearly perfect RAW data. Due to our framework's inherent reversibility, we can reconstruct realistic RAW data instead of synthesizing RAW data from sRGB images without any memory overhead. We also integrate…
Deep Convolutional Neural Networks (CNNs) i.e. Residual Networks (ResNets) have been used successfully for many computer vision tasks, but are difficult to scale to 3D volumetric medical data. Memory is increasingly often the bottleneck when training 3D Convolutional Neural Networks (CNNs). Recently, invertible neural networks have been applied to significantly reduce activation memory footprint when training neural networks with backpropagation thanks to the invertible functions that allow retrieving its input from its output without storing intermediate activations in memory to perform the b…
Yale University（耶鲁大学）；United Imaging Intelligence（联影智能）；University of Georgia（佐治亚大学）
High-quality MRI reconstruction plays a critical role in clinical applications. Deep learning-based methods have achieved promising results on MRI reconstruction. However, most state-of-the-art methods were designed to optimize the evaluation metrics commonly used for natural images, such as PSNR and SSIM, whereas the visual quality is not primarily pursued. Compared to the fully-sampled images, the reconstructed images are often blurry, where high-frequency features might not be sharp enough for confident clinical diagnosis. To this end, we propose an invertible sharpening network (InvSharpNe…
The Chinese University of Hong Kong（香港中文大学）；Caritas Institute of Higher Education（明爱专上学院）
Although digital cameras can acquire high-dynamic range (HDR) images, the captured HDR information are mostly quantized to low-dynamic range (LDR) images for display compatibility and compact storage. In this paper, we propose an invertible tone mapping method that converts the multi-exposure HDR to a true LDR (8-bit per color channel) and reserves the capability to accurately restore the original HDR from this {\em invertible LDR}. Our invertible LDR can mimic the appearance of a user-selected tone mapping style. It can be shared over any existing social network platforms that may re-encode o…
University of Science and Technology of China（中国科学技术大学）；Fudan University（复旦大学）
Deep learning has achieved enormous success in various industrial applications. Companies do not want their valuable data to be stolen by malicious employees to train pirated models. Nor do they wish the data analyzed by the competitors after using them online. We propose a novel solution for dataset protection in this scenario by robustly and reversibly transform the images into adversarial images. We develop a reversible adversarial example generator (RAEG) that introduces slight changes to the images to fool traditional classification models. Even though malicious attacks train pirated mode…
Carnegie Mellon University（卡内基梅隆大学）；University of Southern California（南加州大学）；University of California, Berkeley（加利福尼亚大学伯克利分校）
In this work, we propose a new generative model that is capable of automatically decoupling global and local representations of images in an entirely unsupervised setting, by embedding a generative flow in the VAE framework to model the decoder. Specifically, the proposed model utilizes the variational auto-encoding framework to learn a (low-dimensional) vector of latent variables to capture the global information of an image, which is fed as a conditional input to a flow-based invertible decoder with architecture borrowed from style transfer literature. Experimental results on standard image …
Photoacoustic imaging (PAI) can image high-resolution structures of clinical interest such as vascularity in cancerous tumor monitoring. When imaging human subjects, geometric restrictions force limited-view data retrieval causing imaging artifacts. Iterative physical model based approaches reduce artifacts but require prohibitively time consuming PDE solves. Machine learning (ML) has accelerated PAI by combining physical models and learned networks. However, the depth and overall power of ML methods is limited by memory intensive training. We propose using invertible neural networks (INNs) to…
Guizhou University（贵州大学）；Harbin Medical University Cancer Hospital（哈尔滨医科大学附属肿瘤医院）；Beijing Jiaotong University（北京交通大学）；Shanghai Jiao Tong University（上海交通大学）；University Lyon, INSA Lyon, CNRS, Inserm, IRP Metislab CREATIS UMR5220, U1206（里昂大学、里昂国立应用科学学院、法国国家科学研究中心、法国国家健康与医学研究院、IRP Metislab CREATIS UMR5220、U1206）
In vivo diffusion tensor imaging (DTI) is a promising technique to investigate noninvasively the fiber structures of the in vivo human heart. However, signal loss due to motions remains a persistent problem in in vivo cardiac DTI. We propose a novel motion-compensation method for investigating in vivo myocardium structures in DTI with free-breathing acquisitions. The method is based on an invertible Wavelet Scattering achieved by means of Convolutional Neural Network (WSCNN). It consists of first extracting translation-invariant wavelet scattering features from DW images acquired at different …
Image denoising aims to restore a clean image from an observed noisy image. The model-based image denoising approaches can achieve good generalization ability over different noise levels and are with high interpretability. Learning-based approaches are able to achieve better results, but usually with weaker generalization ability and interpretability. In this paper, we propose a wavelet-inspired invertible network (WINNet) to combine the merits of the wavelet-based approaches and learningbased approaches. The proposed WINNet consists of K-scale of lifting inspired invertible neural networks (L…
The unsupervised image-to-image translation aims at finding a mapping between the source ($A$) and target ($B$) image domains, where in many applications aligned image pairs are not available at training. This is an ill-posed learning problem since it requires inferring the joint probability distribution from marginals. Joint learning of coupled mappings $F_{AB}: A \rightarrow B$ and $F_{BA}: B \rightarrow A$ is commonly used by the state-of-the-art methods, like CycleGAN [Zhu et al., 2017], to learn this translation by introducing cycle consistency requirement to the learning problem, i.e. $F…
Recently pretrained generative models have shown promising results for subsampled Fourier Ptychography (FP) in terms of quality of reconstruction for extremely low sampling rate and high noise. However, one of the significant drawbacks of these pretrained generative priors is their limited representation capabilities. Moreover, training these generative models requires access to a large number of fully-observed clean samples of a particular class of images like faces or digits that is prohibitive to obtain in the context of FP. In this paper, we propose to leverage the power of pretrained inve…
There remains an important need for the development of image reconstruction methods that can produce diagnostically useful images from undersampled measurements. In magnetic resonance imaging (MRI), for example, such methods can facilitate reductions in data-acquisition times. Deep learning-based methods hold potential for learning object priors or constraints that can serve to mitigate the effects of data-incompleteness on image reconstruction. One line of emerging research involves formulating an optimization-based reconstruction method in the latent space of a generative deep neural network…
Korea Advanced Institute of Science and Technology (KAIST)（韩国科学技术院）
Recently, CycleGAN was shown to provide high-performance, ultra-fast denoising for low-dose X-ray computed tomography (CT) without the need for a paired training dataset. Although this was possible thanks to cycle consistency, CycleGAN requires two generators and two discriminators to enforce cycle consistency, demanding significant GPU resources and technical skills for training. A recent proposal of tunable CycleGAN with Adaptive Instance Normalization (AdaIN) alleviates the problem in part by using a single generator. However, two discriminators and an additional AdaIN code generator are st…
National University of Singapore（新加坡国立大学）；National University of Singapore Suzhou Research Institute（新加坡国立大学苏州研究院）
Domain Generalization is a challenging topic in computer vision, especially in Gastrointestinal Endoscopy image analysis. Due to several device limitations and ethical reasons, current open-source datasets are typically collected on a limited number of patients using the same brand of sensors. Different brands of devices and individual differences will significantly affect the model's generalizability. Therefore, to address the generalization problem in GI(Gastrointestinal) endoscopy, we propose a multi-domain GI dataset and a light, plug-in block called InvNorm(Invertible Normalization), whic…
