The recurrent structure is a prevalent framework for the task of video super-resolution, which models the temporal dependency between frames via hidden states. When applied to real-world scenarios with unknown and complex degradations, hidden states tend to contain unpleasant artifacts and propagate them to restored frames. In this circumstance, our analyses show that such artifacts can be largely alleviated when the hidden state is replaced with a cleaner counterpart. Based on the observations, we propose a Hidden State Attention (HSA) module to mitigate artifacts in real-world video super-resolution. Specifically, we first adopt various cheap filters to produce a hidden state pool. For example, Gaussian blur filters are for smoothing artifacts while sharpening filters are for enhancing details. To aggregate a new hidden state that contains fewer artifacts from the hidden state pool, we devise a Selective Cross Attention (SCA) module, in which the attention between input features and each hidden state is calculated. Equipped with HSA, our proposed method, namely FastRealVSR, is able to achieve 2x speedup while obtaining better performance than Real-BasicVSR. Codes will be available at https://github.com/TencentARC/FastRealVSR
Video super-resolution (VSR) aiming to reconstruct a high-resolution (HR) video from its low-resolution (LR) counterpart has made tremendous progress in recent years. However, it remains challenging to deploy existing VSR methods to real-world data with complex degradations. On the one hand, there are few well-aligned real-world VSR datasets, especially with large super-resolution scale factors, which limits the development of real-world VSR tasks. On the other hand, alignment algorithms in existing VSR methods perform poorly for real-world videos, leading to unsatisfactory results. As an attempt to address the aforementioned issues, we build a real-world 4 VSR dataset, namely MVSR4$\times$, where low- and high-resolution videos are captured with different focal length lenses of a smartphone, respectively. Moreover, we propose an effective alignment method for real-world VSR, namely EAVSR. EAVSR takes the proposed multi-layer adaptive spatial transform network (MultiAdaSTN) to refine the offsets provided by the pre-trained optical flow estimation network. Experimental results on RealVSR and MVSR4$\times$ datasets show the effectiveness and practicality of our method, and we achieve state-of-the-art performance in real-world VSR task. The dataset and code will be publicly available.
Video super-resolution (VSR) is a task that aims to reconstruct high-resolution (HR) frames from the low-resolution (LR) reference frame and multiple neighboring frames. The vital operation is to utilize the relative misaligned frames for the current frame reconstruction and preserve the consistency of the results. Existing methods generally explore information propagation and frame alignment to improve the performance of VSR. However, few studies focus on the temporal consistency of inter-frames. In this paper, we propose a Temporal Consistency learning Network (TCNet) for VSR in an end-to-end manner, to enhance the consistency of the reconstructed videos. A spatio-temporal stability module is designed to learn the self-alignment from inter-frames. Especially, the correlative matching is employed to exploit the spatial dependency from each frame to maintain structural stability. Moreover, a self-attention mechanism is utilized to learn the temporal correspondence to implement an adaptive warping operation for temporal consistency among multi-frames. Besides, a hybrid recurrent architecture is designed to leverage short-term and long-term information. We further present a progressive fusion module to perform a multistage fusion of spatio-temporal features. And the final reconstructed frames are refined by these fused features. Objective and subjective results of various experiments demonstrate that TCNet has superior performance on different benchmark datasets, compared to several state-of-the-art methods.
Succinct representation of complex signals using coordinate-based neural representations (CNRs) has seen great progress, and several recent efforts focus on extending them for handling videos. Here, the main challenge is how to (a) alleviate a compute-inefficiency in training CNRs to (b) achieve high-quality video encoding while (c) maintaining the parameter-efficiency. To meet all requirements (a), (b), and (c) simultaneously, we propose neural video representations with learnable positional features (NVP), a novel CNR by introducing "learnable positional features" that effectively amortize a video as latent codes. Specifically, we first present a CNR architecture based on designing 2D latent keyframes to learn the common video contents across each spatio-temporal axis, which dramatically improves all of those three requirements. Then, we propose to utilize existing powerful image and video codecs as a compute-/memory-efficient compression procedure of latent codes. We demonstrate the superiority of NVP on the popular UVG benchmark; compared with prior arts, NVP not only trains 2 times faster (less than 5 minutes) but also exceeds their encoding quality as 34.07$\rightarrow$34.57 (measured with the PSNR metric), even using $>$8 times fewer parameters. We also show intriguing properties of NVP, e.g., video inpainting, video frame interpolation, etc.
In recent years, real image super-resolution (SR) has achieved promising results due to the development of SR datasets and corresponding real SR methods. In contrast, the field of real video SR is lagging behind, especially for real raw videos. Considering the superiority of raw image SR over sRGB image SR, we construct a real-world raw video SR (Real-RawVSR) dataset and propose a corresponding SR method. We utilize two DSLR cameras and a beam-splitter to simultaneously capture low-resolution (LR) and high-resolution (HR) raw videos with 2x, 3x, and 4x magnifications. There are 450 video pairs in our dataset, with scenes varying from indoor to outdoor, and motions including camera and object movements. To our knowledge, this is the first real-world raw VSR dataset. Since the raw video is characterized by the Bayer pattern, we propose a two-branch network, which deals with both the packed RGGB sequence and the original Bayer pattern sequence, and the two branches are complementary to each other. After going through the proposed co-alignment, interaction, fusion, and reconstruction modules, we generate the corresponding HR sRGB sequence. Experimental results demonstrate that the proposed method outperforms benchmark real and synthetic video SR methods with either raw or sRGB inputs. Our code and dataset are available at https://github.com/zmzhang1998/Real-RawVSR.
Video super-resolution (VSR) is the task of restoring high-resolution frames from a sequence of low-resolution inputs. Different from single image super-resolution, VSR can utilize frames' temporal information to reconstruct results with more details. Recently, with the rapid development of convolution neural networks (CNN), the VSR task has drawn increasing attention and many CNN-based methods have achieved remarkable results. However, only a few VSR approaches can be applied to real-world mobile devices due to the computational resources and runtime limitations. In this paper, we propose a \textit{Sliding Window based Recurrent Network} (SWRN) which can be real-time inference while still achieving superior performance. Specifically, we notice that video frames should have both spatial and temporal relations that can help to recover details, and the key point is how to extract and aggregate information. Address it, we input three neighboring frames and utilize a hidden state to recurrently store and update the important temporal information. Our experiment on REDS dataset shows that the proposed method can be well adapted to mobile devices and produce visually pleasant results.
Compressed video super-resolution (VSR) aims to restore high-resolution frames from compressed low-resolution counterparts. Most recent VSR approaches often enhance an input frame by borrowing relevant textures from neighboring video frames. Although some progress has been made, there are grand challenges to effectively extract and transfer high-quality textures from compressed videos where most frames are usually highly degraded. In this paper, we propose a novel Frequency-Transformer for compressed video super-resolution (FTVSR) that conducts self-attention over a joint space-time-frequency domain. First, we divide a video frame into patches, and transform each patch into DCT spectral maps in which each channel represents a frequency band. Such a design enables a fine-grained level self-attention on each frequency band, so that real visual texture can be distinguished from artifacts, and further utilized for video frame restoration. Second, we study different self-attention schemes, and discover that a divided attention which conducts a joint space-frequency attention before applying temporal attention on each frequency band, leads to the best video enhancement quality. Experimental results on two widely-used video super-resolution benchmarks show that FTVSR outperforms state-of-the-art approaches on both uncompressed and compressed videos with clear visual margins. Code is available at https://github.com/researchmm/FTVSR.
We present NeuriCam, a novel deep learning-based system to achieve video capture from low-power dual-mode IoT camera systems. Our idea is to design a dual-mode camera system where the first mode is low-power (1.1 mW) but only outputs grey-scale, low resolution, and noisy video and the second mode consumes much higher power (100 mW) but outputs color and higher resolution images. To reduce total energy consumption, we heavily duty cycle the high power mode to output an image only once every second. The data for this camera system is then wirelessly sent to a nearby plugged-in gateway, where we run our real-time neural network decoder to reconstruct a higher-resolution color video. To achieve this, we introduce an attention feature filter mechanism that assigns different weights to different features, based on the correlation between the feature map and the contents of the input frame at each spatial location. We design a wireless hardware prototype using off-the-shelf cameras and address practical issues including packet loss and perspective mismatch. Our evaluations show that our dual-camera approach reduces energy consumption by 7x compared to existing systems. Further, our model achieves an average greyscale PSNR gain of 3.7 dB over prior single and dual-camera video super-resolution methods and 5.6 dB RGB gain over prior color propagation methods. Open-source code: https://github.com/vb000/NeuriCam.
In this paper, we study a practical space-time video super-resolution (STVSR) problem which aims at generating a high-framerate high-resolution sharp video from a low-framerate low-resolution blurry video. Such problem often occurs when recording a fast dynamic event with a low-framerate and low-resolution camera, and the captured video would suffer from three typical issues: i) motion blur occurs due to object/camera motions during exposure time; ii) motion aliasing is unavoidable when the event temporal frequency exceeds the Nyquist limit of temporal sampling; iii) high-frequency details are lost because of the low spatial sampling rate. These issues can be alleviated by a cascade of three separate sub-tasks, including video deblurring, frame interpolation, and super-resolution, which, however, would fail to capture the spatial and temporal correlations among video sequences. To address this, we propose an interpretable STVSR framework by leveraging both model-based and learning-based methods. Specifically, we formulate STVSR as a joint video deblurring, frame interpolation, and super-resolution problem, and solve it as two sub-problems in an alternate way. For the first sub-problem, we derive an interpretable analytical solution and use it as a Fourier data transform layer. Then, we propose a recurrent video enhancement layer for the second sub-problem to further recover high-frequency details. Extensive experiments demonstrate the superiority of our method in terms of quantitative metrics and visual quality.
The target of space-time video super-resolution (STVSR) is to increase both the frame rate (also referred to as the temporal resolution) and the spatial resolution of a given video. Recent approaches solve STVSR using end-to-end deep neural networks. A popular solution is to first increase the frame rate of the video; then perform feature refinement among different frame features; and last increase the spatial resolutions of these features. The temporal correlation among features of different frames is carefully exploited in this process. The spatial correlation among features of different (spatial) resolutions, despite being also very important, is however not emphasized. In this paper, we propose a spatial-temporal feature interaction network to enhance STVSR by exploiting both spatial and temporal correlations among features of different frames and spatial resolutions. Specifically, the spatial-temporal frame interpolation module is introduced to interpolate low- and high-resolution intermediate frame features simultaneously and interactively. The spatial-temporal local and global refinement modules are respectively deployed afterwards to exploit the spatial-temporal correlation among different features for their refinement. Finally, a novel motion consistency loss is employed to enhance the motion continuity among reconstructed frames. We conduct experiments on three standard benchmarks, Vid4, Vimeo-90K and Adobe240, and the results demonstrate that our method improves the state of the art methods by a considerable margin. Our codes will be available at https://github.com/yuezijie/STINet-Space-time-Video-Super-resolution.
The success of existing video super-resolution (VSR) algorithms stems mainly exploiting the temporal information from the neighboring frames. However, none of these methods have discussed the influence of the temporal redundancy in the patches with stationary objects and background and usually use all the information in the adjacent frames without any discrimination. In this paper, we observe that the temporal redundancy will bring adverse effect to the information propagation,which limits the performance of the most existing VSR methods. Motivated by this observation, we aim to improve existing VSR algorithms by handling the temporal redundancy patches in an optimized manner. We develop two simple yet effective plug and play methods to improve the performance of existing local and non-local propagation-based VSR algorithms on widely-used public videos. For more comprehensive evaluating the robustness and performance of existing VSR algorithms, we also collect a new dataset which contains a variety of public videos as testing set. Extensive evaluations show that the proposed methods can significantly improve the performance of existing VSR methods on the collected videos from wild scenarios while maintain their performance on existing commonly used datasets. The code is available at https://github.com/HYHsimon/Boosted-VSR.
The alignment of adjacent frames is considered an essential operation in video super-resolution (VSR). Advanced VSR models, including the latest VSR Transformers, are generally equipped with well-designed alignment modules. However, the progress of the self-attention mechanism may violate this common sense. In this paper, we rethink the role of alignment in VSR Transformers and make several counter-intuitive observations. Our experiments show that: (i) VSR Transformers can directly utilize multi-frame information from unaligned videos, and (ii) existing alignment methods are sometimes harmful to VSR Transformers. These observations indicate that we can further improve the performance of VSR Transformers simply by removing the alignment module and adopting a larger attention window. Nevertheless, such designs will dramatically increase the computational burden, and cannot deal with large motions. Therefore, we propose a new and efficient alignment method called patch alignment, which aligns image patches instead of pixels. VSR Transformers equipped with patch alignment could demonstrate state-of-the-art performance on multiple benchmarks. Our work provides valuable insights on how multi-frame information is used in VSR and how to select alignment methods for different networks/datasets. Codes and models will be released at https://github.com/XPixelGroup/RethinkVSRAlignment.
Despite the quality improvement brought by the recent methods, video super-resolution (SR) is still very challenging, especially for videos that are low-light and noisy. The current best solution is to subsequently employ best models of video SR, denoising, and illumination enhancement, but doing so often lowers the image quality, due to the inconsistency between the models. This paper presents a new parametric representation called the Deep Parametric 3D Filters (DP3DF), which incorporates local spatiotemporal information to enable simultaneous denoising, illumination enhancement, and SR efficiently in a single encoder-and-decoder network. Also, a dynamic residual frame is jointly learned with the DP3DF via a shared backbone to further boost the SR quality. We performed extensive experiments, including a large-scale user study, to show our method's effectiveness. Our method consistently surpasses the best state-of-the-art methods on all the challenging real datasets with top PSNR and user ratings, yet having a very fast run time.
The high computational costs of video super-resolution (VSR) models hinder their deployment on resource-limited devices, (e.g., smartphones and drones). Existing VSR models contain considerable redundant filters, which drag down the inference efficiency. To prune these unimportant filters, we develop a structured pruning scheme called Structured Sparsity Learning (SSL) according to the properties of VSR. In SSL, we design pruning schemes for several key components in VSR models, including residual blocks, recurrent networks, and upsampling networks. Specifically, we develop a Residual Sparsity Connection (RSC) scheme for residual blocks of recurrent networks to liberate pruning restrictions and preserve the restoration information. For upsampling networks, we design a pixel-shuffle pruning scheme to guarantee the accuracy of feature channel-space conversion. In addition, we observe that pruning error would be amplified as the hidden states propagate along with recurrent networks. To alleviate the issue, we design Temporal Finetuning (TF). Extensive experiments show that SSL can significantly outperform recent methods quantitatively and qualitatively.
This paper studies the problem of real-world video super-resolution (VSR) for animation videos, and reveals three key improvements for practical animation VSR. First, recent real-world super-resolution approaches typically rely on degradation simulation using basic operators without any learning capability, such as blur, noise, and compression. In this work, we propose to learn such basic operators from real low-quality animation videos, and incorporate the learned ones into the degradation generation pipeline. Such neural-network-based basic operators could help to better capture the distribution of real degradations. Second, a large-scale high-quality animation video dataset, AVC, is built to facilitate comprehensive training and evaluations for animation VSR. Third, we further investigate an efficient multi-scale network structure. It takes advantage of the efficiency of unidirectional recurrent networks and the effectiveness of sliding-window-based methods. Thanks to the above delicate designs, our method, AnimeSR, is capable of restoring real-world low-quality animation videos effectively and efficiently, achieving superior performance to previous state-of-the-art methods. Codes and models are available at https://github.com/TencentARC/AnimeSR.
Videos typically record the streaming and continuous visual data as discrete consecutive frames. Since the storage cost is expensive for videos of high fidelity, most of them are stored in a relatively low resolution and frame rate. Recent works of Space-Time Video Super-Resolution (STVSR) are developed to incorporate temporal interpolation and spatial super-resolution in a unified framework. However, most of them only support a fixed up-sampling scale, which limits their flexibility and applications. In this work, instead of following the discrete representations, we propose Video Implicit Neural Representation (VideoINR), and we show its applications for STVSR. The learned implicit neural representation can be decoded to videos of arbitrary spatial resolution and frame rate. We show that VideoINR achieves competitive performances with state-of-the-art STVSR methods on common up-sampling scales and significantly outperforms prior works on continuous and out-of-training-distribution scales. Our project page is at http://zeyuan-chen.com/VideoINR/ .
Video restoration aims at restoring multiple high-quality frames from multiple low-quality frames. Existing video restoration methods generally fall into two extreme cases, i.e., they either restore all frames in parallel or restore the video frame by frame in a recurrent way, which would result in different merits and drawbacks. Typically, the former has the advantage of temporal information fusion. However, it suffers from large model size and intensive memory consumption; the latter has a relatively small model size as it shares parameters across frames; however, it lacks long-range dependency modeling ability and parallelizability. In this paper, we attempt to integrate the advantages of the two cases by proposing a recurrent video restoration transformer, namely RVRT. RVRT processes local neighboring frames in parallel within a globally recurrent framework which can achieve a good trade-off between model size, effectiveness, and efficiency. Specifically, RVRT divides the video into multiple clips and uses the previously inferred clip feature to estimate the subsequent clip feature. Within each clip, different frame features are jointly updated with implicit feature aggregation. Across different clips, the guided deformable attention is designed for clip-to-clip alignment, which predicts multiple relevant locations from the whole inferred clip and aggregates their features by the attention mechanism. Extensive experiments on video super-resolution, deblurring, and denoising show that the proposed RVRT achieves state-of-the-art performance on benchmark datasets with balanced model size, testing memory and runtime.
Upscaled video detection is a helpful tool in multimedia forensics, but it is a challenging task that involves various upscaling and compression algorithms. There are many resolution-enhancement methods, including interpolation and deep-learning-based super-resolution, and they leave unique traces. In this work, we propose a new upscaled-resolution-detection method based on learning of visual representations using contrastive and cross-entropy losses. To explain how the method detects videos, we systematically review the major components of our framework - in particular, we show that most data-augmentation approaches hinder the learning of the method. Through extensive experiments on various datasets, we demonstrate that our method effectively detects upscaling even in compressed videos and outperforms the state-of-the-art alternatives. The code and models are publicly available at https://github.com/msu-video-group/SRDM
How to properly model the inter-frame relation within the video sequence is an important but unsolved challenge for video restoration (VR). In this work, we propose an unsupervised flow-aligned sequence-to-sequence model (S2SVR) to address this problem. On the one hand, the sequence-to-sequence model, which has proven capable of sequence modeling in the field of natural language processing, is explored for the first time in VR. Optimized serialization modeling shows potential in capturing long-range dependencies among frames. On the other hand, we equip the sequence-to-sequence model with an unsupervised optical flow estimator to maximize its potential. The flow estimator is trained with our proposed unsupervised distillation loss, which can alleviate the data discrepancy and inaccurate degraded optical flow issues of previous flow-based methods. With reliable optical flow, we can establish accurate correspondence among multiple frames, narrowing the domain difference between 1D language and 2D misaligned frames and improving the potential of the sequence-to-sequence model. S2SVR shows superior performance in multiple VR tasks, including video deblurring, video super-resolution, and compressed video quality enhancement. Code and models are publicly available at https://github.com/linjing7/VR-Baseline
Document-level relation extraction (RE) aims at extracting relations among entities expressed across multiple sentences, which can be viewed as a multi-label classification problem. In a typical document, most entity pairs do not express any pre-defined relation and are labeled as "none" or "no relation". For good document-level RE performance, it is crucial to distinguish such none class instances (entity pairs) from those of pre-defined classes (relations). However, most existing methods only estimate the probability of pre-defined relations independently without considering the probability of "no relation". This ignores the context of entity pairs and the label correlations between the none class and pre-defined classes, leading to sub-optimal predictions. To address this problem, we propose a new multi-label loss that encourages large margins of label confidence scores between each pre-defined class and the none class, which enables captured label correlations and context-dependent thresholding for label prediction. To gain further robustness against positive-negative imbalance and mislabeled data that could appear in real-world RE datasets, we propose a margin regularization and a margin shifting technique. Experimental results demonstrate that our method significantly outperforms existing multi-label losses for document-level RE and works well in other multi-label tasks such as emotion classification when none class instances are available for training.