光场视频插帧方法（时空超分/时域超分）

… suffering from low frame rates when providing high spatial-angular sampling. To address the above challenges, this paper proposes a new LFV frame interpolation method based on …

In this paper, we address the problem of temporal interpolation of sparsely sampled video light fields using dense scene flows. Given light fields at two time instants, the goal is to interpolate an intermediate light field to form a spatially, angularly and temporally coherent light field video sequence. We first compute angularly coherent bidirectional scene flows between the two input light fields. We then use the optical flows and the two light fields as inputs to a convolutional neural network that synthesizes independently the views of the light field at an intermediate time. In order to measure the angular consistency of a light field, we propose a new metric based on epipolar geometry. Experimental results show that the proposed method produces light fields that are angularly coherent while keeping similar temporal and spatial consistency as state-of-the-art video frame interpolation methods.

Light field cameras have many advantages over traditional cameras, as they allow the user to change various camera settings after capture. However, capturing light fields requires a huge bandwidth to record the data: a modern light field camera can only take three images per second. This prevents current consumer light field cameras from capturing light field videos. Temporal interpolation at such extreme scale (10x, from 3 fps to 30 fps) is infeasible as too much information will be entirely missing between adjacent frames. Instead, we develop a hybrid imaging system, adding another standard video camera to capture the temporal information. Given a 3 fps light field sequence and a standard 30 fps 2D video, our system can then generate a full light field video at 30 fps. We adopt a learning-based approach, which can be decomposed into two steps: spatio-temporal flow estimation and appearance estimation. The flow estimation propagates the angular information from the light field sequence to the 2D video, so we can warp input images to the target view. The appearance estimation then combines these warped images to output the final pixels. The whole process is trained end-to-end using convolutional neural networks. Experimental results demonstrate that our algorithm outperforms current video interpolation methods, enabling consumer light field videography, and making applications such as refocusing and parallax view generation achievable on videos for the first time.

Light field (LF) cameras suffer from a fundamental trade-off between spatial and angular resolutions. Additionally, due to the significant amount of data that needs to be recorded, the Lytro ILLUM, a modern LF camera, can only capture three frames per second. In this paper, we consider space-time super-resolution (SR) for LF videos, aiming at generating high-resolution and high-frame-rate LF videos from low-resolution and low-frame-rate observations. Extending existing space-time video SR methods to this task directly will meet two key challenges: 1) how to re-organize sub-aperture images (SAIs) efficiently and effectively given highly redundant LF videos, and 2) how to aggregate complementary information between multiple SAIs and frames considering the coherence in LF videos. To address the above challenges, we propose a novel framework for space-time super-resolving LF videos for the first time. First, we propose a novel Multi-Scale Dilated SAI Re-organization strategy for re-organizing SAIs into auxiliary view stacks with decreasing resolution as the Chebyshev distance in the angular dimension increases. In particular, the auxiliary view stack with original resolution preserves essential visual details, while the down-scaled view stacks capture long-range contextual information. Second, we propose the Multi-Scale Aggregated Feature extractor and the Angular-Assisted Feature Interpolation module to utilize and aggregate information from the spatial, angular, and temporal dimensions in LF videos. The former aggregates similar contents from different SAIs and frames for subsequent reconstruction in a disparity-free manner at the feature level, whereas the latter interpolates intermediate frames temporally by implicitly aggregating geometric information. Compared to other potential approaches, experimental results demonstrate that the reconstructed LF videos generated by our framework achieve higher reconstruction quality and better preserve the LF parallax structure and temporal consistency. The implementation code is available at https://github.com/zeyuxiao1997/LFSTVSR.

A factored display emits full-parallax dense-view light fields for a glasses-free 3D experience without sacrificing the spatial resolution of a liquid-crystal display (LCD). For static light fields, it achieves high-quality reconstruction by applying frame-based low-rank factorization to time-multiplexed sub-frame contents of stacked LCDs. However, for light field videos such frame-based factorization could introduce reconstruction artifacts and visual flickers and further cause human discomfort. The artifacts mainly come from incomplete constraints for the emitted light fields that are actually perceived in continuous time, instead of discrete frames. In particular, the perceived light fields are related to the persistence-of-vision (POV) effect of human eyes and the refresh rates of LCD displays, which is not well explored in previous work. In this work, we introduce a light-field video factorization framework—temporal fusion (TF)—to resolve these issues. To begin with, we explicitly formulate the continuous-time POV effect into a global factorization objective functional to eliminate visual flickers and enhance image quality. We further show that this optimization problem can be solved by sequence-level iterative updates on LCD sub-frames. Then, to tackle the enormous requirement of memory access for the sequence-level processing flow, we devise an efficient cuboid-wise factorization algorithm which enables practical GPU implementation. We also devise another lightweight causal framework, TF-C, for supporting low-latency applications. Finally, extensive experiments are performed to verify the effectiveness. Compared to the plain frame-based factorization, TF/TF-C can improve temporal consistency by reducing flicker values by 85%/91% and enhance reconstruction quality by increasing PSNR values by 5.0dB/3.7dB. In addition, we present a prototype dual-layer factored display, which was built with two 240-Hz high-refresh-rate LCDs, to demonstrate the visual quality for real-life applications.

… Inter-frame modeling is pivotal in generating intermediate frames for video frame interpolation (VFI)… Finally, we use the extracted inter-frame features to estimate motion and generate the …

With the advancement of AIGC, video frame interpolation (VFI) has become a crucial component in existing video generation frameworks, attracting widespread research interest. For the VFI task, the motion estimation between neighboring frames plays a crucial role in avoiding motion ambiguity. However, existing VFI methods always struggle to accurately predict the motion information between consecutive frames, and this imprecise estimation leads to blurred and visually incoherent interpolated frames. In this paper, we propose a novel diffusion framework, Motion-Aware latent Diffusion models (MADiff), which is specifically designed for the VFI task. By incorporating motion priors between the conditional neighboring frames with the target interpolated frame predicted throughout the diffusion sampling procedure, MADiff progressively refines the intermediate outcomes, culminating in generating both visually smooth and realistic results. Extensive experiments conducted on benchmark datasets demonstrate that our method achieves state-of-the-art performance significantly outperforming existing approaches, especially under challenging scenarios involving dynamic textures with complex motion.

Effectively extracting inter-frame motion and appearance information is important for video frame interpolation (VFI). Previous works either extract both types of information in a mixed way or devise separate modules for each type of information, which lead to representation ambiguity and low efficiency. In this paper, we propose a new module to explicitly extract motion and appearance information via a unified operation. Specifically, we rethink the information process in inter-frame attention and reuse its attention map for both appearance feature enhancement and motion information extraction. Furthermore, for efficient VFI, our proposed module could be seamlessly integrated into a hybrid CNN and Transformer architecture. This hybrid pipeline can alleviate the computational complexity of inter-frame attention as well as preserve detailed low-level structure information. Experimental results demonstrate that, for both fixed- and arbitrary-timestep interpolation, our method achieves state-of-the-art performance on various datasets. Meanwhile, our approach enjoys a lighter computation overhead over models with close performance. The source code and models are available at https://github.com/MCG-NJU/EMA-VFI.

Conventional photography can only provide a two-dimensional image of the scene, whereas emerging imaging modalities such as light field enable the representation of higher dimensional visual information by capturing light rays from different directions. Light fields provide immersive experiences, a sense of presence in the scene, and can enhance different vision tasks. Hence, research into light field processing methods has become increasingly popular. It does, however, come at the cost of higher data volume and computational complexity. With the growing deployment of machine-learning and deep architectures in image processing applications, a paradigm shift toward learning-based approaches has also been observed in the design of light field processing methods. Various learning-based approaches are developed to process the high volume of light field data efficiently for different vision tasks while improving performance. Taking into account the diversity of light field vision tasks and the deployed learning-based frameworks, it is necessary to survey the scattered learning-based works in the domain to gain insight into the current trends and challenges. This paper aims to review the existing learning-based solutions for light field imaging and to summarize the most promising frameworks. Moreover, evaluation methods and available light field datasets are highlighted. Lastly, the review concludes with a brief outlook for future research directions.

… for Light-Field video. The proposed algorithms are specially tailored to process sparse and widebaseline multi-view videos … In this paper we focus on camera arrays as a video LightField …

In this paper, we present a full view optical flow estimation method for plenoptic imaging. Our method employs the structure delivered by the four-dimensional light field over multiple views making use of superpixels. These superpixels are four dimensional in nature and can be used to represent the objects in the scene as a set of slanted-planes in three-dimensional space so as to recover a piecewise rigid depth estimate. Taking advantage of these superpixels and the corresponding slanted planes, we recover the optical flow and depth maps by using a two-step optimization scheme where the flow is propagated from the central view to the other views in the imagery. We illustrate the utility of our method for depth and flow estimation making use of a dataset of synthetically generated image sequences and real-world imagery captured using a Lytro Illum camera. We also compare our results with those yielded by a number of alternatives elsewhere in the literature.

… based on multi-resolution optical flow [BWS05, YWA10, LYT11… -resolution processing to multi-view data has been proposed … multi-view depth reconstruction method for video light fields. …

… In these studies regularization is imposed on the disparity and optical flow (2D formulation), while our assumptions refer directly to the 3D unknowns. Nor were these methods extended …

Light-field imaging is appealing to the mobile devices market because of its capability for intuitive post-capture processing. Acquiring light field (LF) data with high angular, spatial and temporal resolution poses significant challenges, especially with space constraints preventing bulky optics. At the same time, stereo video capture, now available on many consumer devices, can be interpreted as a sparse LF-capture. We explore the application of small baseline stereo videos for reconstructing high fidelity LF videos.We propose a self-supervised learning-based algorithm for LF video reconstruction from stereo video. The self- supervised LF video reconstruction is guided via the geometric information from the individual stereo pairs and the temporal information from the video sequence. LF estimation is further regularized by a low-rank constraint based on layered LF displays. The proposed self-supervised algorithm facilitates advantages such as post-training finetuning on test sequences and variable angular view interpolation and extrapolation. Quantitatively the reconstructed LF videos show higher fidelity than previously proposed unsupervised approaches. We demonstrate our results via LF videos generated from publicly available stereo videos acquired from commercially available stereoscopic cameras. Finally, we demonstrate that our reconstructed LF videos allow applications such as post-capture focus control and region-of-interest (RoI) based focus tracking for videos.

The hardware challenges associated with light-field(LF) imaging has made it difficult for consumers to access its benefits like applications in post-capture focus and aperture control. Learning-based techniques which solve the ill-posed problem of LF reconstruction from sparse (1, 2 or 4) views have significantly reduced the requirement for complex hardware. LF video reconstruction from sparse views poses a special challenge as acquiring ground-truth for training these models is hard. Hence, we propose a self-supervised learning-based algorithm for LF video reconstruction from monocular videos. We use self-supervised geometric, photometric and temporal consistency constraints inspired from a recent self-supervised technique for LF video reconstruction from stereo video. Additionally, we propose three key techniques that are relevant to our monocular video input. We propose an explicit disocclusion handling technique that encourages the network to inpaint disoccluded regions in a LF frame, using information from adjacent input temporal frames. This is crucial for a self-supervised technique as a single input frame does not contain any information about the disoccluded regions. We also propose an adaptive low-rank representation that provides a significant boost in performance by tailoring the representation to each input scene. Finally, we also propose a novel refinement block that is able to exploit the available LF image data using supervised learning to further refine the reconstruction quality. Our qualitative and quantitative analysis demonstrates the significance of each of the proposed building blocks and also the superior results compared to previous state-of-the-art monocular LF reconstruction techniques. We further validate our algorithm by reconstructing LF videos from monocular videos acquired using a commercial GoPro camera.

In this paper, we address the space-time video super-resolution, which aims at generating a high-resolution (HR) slow-motion video from a low-resolution (LR) and low frame rate (LFR) video sequence. A naïve method is to decompose it into two sub-tasks: video frame interpolation (VFI) and video super-resolution (VSR). Nevertheless, temporal interpolation and spatial upscaling are intra-related in this problem. Two-stage approaches cannot fully make use of this natural property. Besides, state-of-the-art VFI or VSR deep networks usually have a large frame reconstruction module in order to obtain high-quality photo-realistic video frames, which makes the two-stage approaches have large models and thus be relatively time-consuming. To overcome the issues, we present a one-stage space-time video super-resolution framework, which can directly reconstruct an HR slow-motion video sequence from an input LR and LFR video. Instead of reconstructing missing LR intermediate frames as VFI models do, we temporally interpolate LR frame features of the missing LR frames capturing local temporal contexts by a feature temporal interpolation module. Extensive experiments on widely used benchmarks demonstrate that the proposed framework not only achieves better qualitative and quantitative performance on both clean and noisy LR frames but also is several times faster than recent state-of-the-art two-stage networks. The source code is released in https://github.com/Mukosame/Zooming-Slow-Mo-CVPR-2020 .

The capability of video super-resolution (VSR) to synthesize high-resolution (HR) video from ideal datasets has been demonstrated in many works. However, applying the VSR model to real-world video with unknown and complex degradation remains a challenging task. First, existing degradation metrics in most VSR methods are not able to effectively simulate real-world noise and blur. On the contrary, simple combinations of classical degradation are used for real-world noise modeling, which led to the VSR model often being violated by out-of-distribution noise. Second, many SR models focus on noise simulation and transfer. Nevertheless, the sampled noise is monotonous and limited. To address the aforementioned problems, we propose a Negatives augmentation strategy for generalized noise modeling in Video Super-Resolution (NegVSR) task. Specifically, we first propose sequential noise generation toward real-world data to extract practical noise sequences. Then, the degeneration domain is widely expanded by negative augmentation to build up various yet challenging real-world noise sets. We further propose the augmented negative guidance loss to learn robust features among augmented negatives effectively. Extensive experiments on real-world datasets (e.g., VideoLQ and FLIR) show that our method outperforms state-of-the-art methods with clear margins, especially in visual quality. Project page is available at: https://negvsr.github.io/.

Video super-resolution plays an important role in surveillance video analysis and ultra-high-definition video display, which has drawn much attention in both the research and industrial communities. Although many deep learning-based VSR methods have been proposed, it is hard to directly compare these methods since the different loss functions and training datasets have a significant impact on the super-resolution results. In this work, we carefully study and compare three temporal modeling methods (2D CNN with early fusion, 3D CNN with slow fusion and Recurrent Neural Network) for video super-resolution. We also propose a novel Recurrent Residual Network (RRN) for efficient video super-resolution, where residual learning is utilized to stabilize the training of RNN and meanwhile to boost the super-resolution performance. Extensive experiments show that the proposed RRN is highly computational efficiency and produces temporal consistent VSR results with finer details than other temporal modeling methods. Besides, the proposed method achieves state-of-the-art results on several widely used benchmarks.

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.

Deep learning-based blind super-resolution (SR) methods have recently achieved unprecedented performance in upscaling frames with unknown degradation. These models are able to accurately estimate the unknown downscaling kernel from a given low-resolution (LR) image in order to leverage the kernel during restoration. Although these approaches have largely been successful, they are predominantly image-based and therefore do not exploit the temporal properties of the kernels across multiple video frames. In this paper, we investigated the temporal properties of the kernels and highlighted its importance in the task of blind video super-resolution. Specifically, we measured the kernel temporal consistency of real-world videos and illustrated how the estimated kernels might change per frame in videos of varying dynamicity of the scene and its objects. With this new insight, we revisited previous popular video SR approaches, and showed that previous assumptions of using a fixed kernel throughout the restoration process can lead to visual artifacts when upscaling real-world videos. In order to counteract this, we tailored existing single-image and video SR techniques to leverage kernel consistency during both kernel estimation and video upscaling processes. Extensive experiments on synthetic and real-world videos show substantial restoration gains quantitatively and qualitatively, achieving the new state-of-the-art in blind video SR and underlining the potential of exploiting kernel temporal consistency.

Optical-flow-based and kernel-based approaches have been extensively explored for temporal compensation in satellite Video Super-Resolution (VSR). However, these techniques are less generalized in large-scale or complex scenarios, especially in satellite videos. In this paper, we propose to exploit the well-defined temporal difference for efficient and effective temporal compensation. To fully utilize the local and global temporal information within frames, we systematically modeled the short-term and long-term temporal discrepancies since we observed that these discrepancies offer distinct and mutually complementary properties. Specifically, we devise a Short-term Temporal Difference Module (S-TDM) to extract local motion representations from RGB difference maps between adjacent frames, which yields more clues for accurate texture representation. To explore the global dependency in the entire frame sequence, a Long-term Temporal Difference Module (L-TDM) is proposed, where the differences between forward and backward segments are incorporated and activated to guide the modulation of the temporal feature, leading to a holistic global compensation. Moreover, we further propose a Difference Compensation Unit (DCU) to enrich the interaction between the spatial distribution of the target frame and temporal compensated results, which helps maintain spatial consistency while refining the features to avoid misalignment. Rigorous objective and subjective evaluations conducted across five mainstream video satellites demonstrate that our method performs favorably against state-of-the-art approaches. Code will be available at https://github.com/XY-boy/LGTD