A promising direction for recovering the lost information in low-resolution headshot images is utilizing a set of high-resolution exemplars from the same identity. Complementary images in the reference set can improve the generated headshot quality across many different views and poses. However, it is challenging to make the best use of multiple exemplars: the quality and alignment of each exemplar cannot be guaranteed. Using low-quality and mismatched images as references will impair the output results. To overcome these issues, we propose the Headshot Image Super-Resolution with Multiple Exemplars network (HIME) method. Compared with previous methods, our network can effectively handle the misalignment between the input and the reference without requiring facial priors and learn the aggregated reference set representation in an end-to-end manner. Furthermore, to reconstruct more detailed facial features, we propose a correlation loss that provides a rich representation of the local texture in a controllable spatial range. Experimental results demonstrate that the proposed framework not only has significantly fewer computation cost than recent exemplar-guided methods but also achieves better qualitative and quantitative performance.

Generative adversarial networks (GANs), e.g., StyleGAN2, play a vital role in various image generation and synthesis tasks, yet their notoriously high computational cost hinders their efficient deployment on edge devices. Directly applying generic compression approaches yields poor results on GANs, which motivates a number of recent GAN compression works. While prior works mainly accelerate conditional GANs, e.g., pix2pix and Cycle-GAN, compressing state-of-the-art unconditional GANs has rarely been explored and is more challenging. In this paper, we propose novel approaches for unconditional GAN compression. We first introduce effective channel pruning and knowledge distillation schemes specialized for unconditional GANs. We then propose a novel content-aware method to guide the processes of both pruning and distillation. With content-awareness, we can effectively prune channels that are unimportant to the contents of interest, e.g., human faces, and focus our distillation on these regions, which significantly enhances the distillation quality. On StyleGAN2 and SN-GAN, we achieve a substantial improvement over the state-of-the-art compression method. Notably, we reduce the FLOPs of StyleGAN2 by 11× with visually negligible image quality loss compared to the full-size model. More interestingly, when applied to various image manipulation tasks, our compressed model forms a smoother and better disentangled latent manifold, making it more effective for image editing.

We introduce a neural network-based method to denoise pairs of images taken in quick succession, with and without a flash, in low-light environments. Our goal is to produce a high-quality rendering of the scene that preserves the color and mood from the ambient illumination of the noisy no-flash image, while recovering surface texture and detail revealed by the flash. Our network outputs a gain map and a field of kernels, the latter obtained by linearly mixing elements of a per-image low-rank kernel basis. We first apply the kernel field to the no-flash image, and then multiply the result with the gain map to create the final output. We show our network effectively learns to produce high-quality images by combining a smoothed out estimate of the scene’s ambient appearance from the no-flash image, with high-frequency albedo details extracted from the flash input. Our experiments show significant improvements over alternative captures without a flash, and baseline denoisers that use flash no-flash pairs. In particular, our method produces images that are both noise-free and contain accurate ambient colors without the sharp shadows or strong specular highlights visible in the flash image.

Image compositing is a task of combining regions from different images to compose a new image. A common use case is background replacement of portrait images. To obtain high quality composites, professionals typically manually perform multiple editing steps such as segmentation, matting and foreground color decontamination, which is very time consuming even with sophisticated photo editing tools. In this paper, we propose a new method which can automatically generate high-quality image compositing with-out any user input. Our method can be trained end-to-end to optimize exploitation of contextual and color information of both foreground and background images, where the com-positing quality is considered in the optimization. Specifically, inspired by Laplacian pyramid blending, a dense-connected multi-stream fusion network is proposed to effectively fuse the information from the foreground and back-ground images at different scales. In addition, we intro-duce a self-taught strategy to progressively train from easy to complex cases to mitigate the lack of training data. Experiments show that the proposed method can automatically generate high-quality composites and outperforms existing methods both qualitatively and quantitatively.

Current methods for active speaker detection focus on modeling audiovisual information from a single speaker. This strategy can be adequate for addressing single-speaker scenarios, but it prevents accurate detection when the task is to identify who of many candidate speakers are talking. This paper introduces the Active Speaker Context, a novel representation that models relationships between multiple speakers over long time horizons. Our new model learns pairwise and temporal relations from a structured ensemble of audiovisual observations. Our experiments show that a structured feature ensemble already benefits active speaker detection performance. We also find that the proposed Active Speaker Context improves the state-of-the-art on the AVA-ActiveSpeaker dataset achieving an mAP of 87.1%. Moreover, ablation studies verify that this result is a direct consequence of our long-term multi-speaker analysis.

We present TDNet, a temporally distributed network designed for fast and accurate video semantic segmentation. We observe that features extracted from a certain high-level layer of a deep CNN can be approximated by composing features extracted from several shallower sub-networks. Leveraging the inherent temporal continuity in videos, we distribute these sub-networks over sequential frames. Therefore, at each time step, we only need to perform a lightweight computation to extract a sub-features group from a single sub-network. The full features used for segmentation are then recomposed by application of a novel attention propagation module that compensates for geometry deformation between frames. A grouped knowledge distillation loss is also introduced to further improve the representation power at both full and sub-feature levels. Experiments on Cityscapes, CamVid, and NYUD-v2 demonstrate that our method achieves state-of-the-art accuracy with significantly faster speed and lower latency.

Bursts of images exhibit significant self-similarity across both time and space. This motivates a representation of the kernels as linear combinations of a small set of basis elements. To this end, we introduce a novel basis prediction network that, given an input burst, predicts a set of global basis kernels --- shared within the image --- and the corresponding mixing coefficients --- which are specific to individual pixels. Compared to state-of-the-art techniques that output a large tensor of per-pixel spatiotemporal kernels, our formulation substantially reduces the dimensionality of the network output. This allows us to effectively exploit comparatively larger denoising kernels, achieving both significant quality improvements (over 1dB PSNR) and faster run-times over state-of-the-art methods.

In this paper, we aim to reduce the footskate artifacts when reconstructing human dynamics from monocular RGB videos. Recent work has made substantial progress in improving the temporal smoothness of the reconstructed motion trajectories. Their results, however, still suffer from severe foot skating and slippage artifacts. To tackle this issue, we present a neural network based detector for localizing ground contact events of human feet and use it to impose a physical constraint for optimization of the whole human dynamics in a video. We present a detailed study on the proposed ground contact detector and demonstrate high-quality human motion reconstruction results in various videos.

We propose a novel multi-stream architecture and training methodology that exploits semantic labels for facial image deblurring. The proposed Uncertainty Guided Multi-Stream Semantic Network (UMSN) processes regions belonging to each semantic class independently and learns to combine their outputs into the final deblurred result. Pixel-wise semantic labels are obtained using a segmentation network. A predicted confidence measure is used during training to guide the network towards the challenging regions of the human face such as the eyes and nose. The entire network is trained in an end-to-end fashion. Comprehensive experiments on three different face datasets demonstrate that the proposed method achieves significant improvements over the recent state-of-the-art face deblurring methods. Code is available at.

We present a fully data-driven method to compute depth from diverse monocular video sequences that contain large amounts of non-rigid objects, e.g., people. In order to learn reconstruction cues for non-rigid scenes, we introduce a new dataset consisting of stereo videos scraped in-the-wild. This dataset has a wide variety of scene types, and features large amounts of nonrigid objects, especially people. From this, we compute disparity maps to be used as supervision to train our approach. We propose a loss function that allows us to generate a depth prediction even with unknown camera intrinsics and stereo baselines in the dataset. We validate the use of large amounts of Internet video by evaluating our method on existing video datasets with depth supervision, including SINTEL, and KITTI, and show that our approach generalizes better to natural scenes.

We present a method to improve the visual realism of low-quality, synthetic images, e.g. OpenGL renderings. Training an unpaired synthetic-to-real translation network in image space is severely under-constrained and produces visible artifacts. Instead, we propose a semi-supervised approach that operates on the disentangled shading and albedo layers of the image. Our two-stage pipeline first learns to predict accurate shading in a supervised fashion using physically-based renderings as targets, and further increases the realism of the textures and shading with an improved CycleGAN network. Extensive evaluations on the SUNCG indoor scene dataset demonstrate that our approach yields more realistic images compared to other state-of-the-art approaches. Furthermore, networks trained on our generated ``real'' images predict more accurate depth and normals than domain adaptation approaches, suggesting that improving the visual realism of the images can be more effective than imposing task-specific losses.

Large scale object detection datasets are constantly increasing their size in terms of the number of classes and annotations count. Yet, the number of object-level categories annotated in detection datasets is an order of magnitude smaller than image-level classification labels. State-of-the art object detection models are trained in a supervised fashion and this limits the number of object classes they can detect. In this paper, we propose a novel weight transfer network (WTN) to effectively and efficiently transfer knowledge from classification network's weights to detection network's weights to allow detection of novel classes without box supervision. We first introduce input and feature normalization schemes to curb the under-fitting during training of a vanilla WTN. We then propose autoencoder-WTN (AE-WTN) which uses reconstruction loss to preserve classification network's information over all classes in the target latent space to ensure generalization to novel classes. Compared to vanilla WTN, AE-WTN obtains absolute performance gains of 6% on two Open Images evaluation sets with 500 seen and 57 novel classes respectively, and 25% on a Visual Genome evaluation set with 200 novel classes.

Most recent semantic segmentation methods train deep convolutional neural networks with fully annotated masks requiring pixel-accuracy for good quality training. Common weakly-supervised approaches generate full masks from partial input (e.g. scribbles or seeds) using standard interactive segmentation methods as preprocessing. But, errors in such masks result in poorer training since standard loss functions (e.g. cross-entropy) do not distinguish seeds from potentially mislabeled other pixels. Inspired by the general ideas in semi-supervised learning, we address these problems via a new principled loss function evaluating network output with criteria standard in "shallow" segmentation, e.g. normalized cut. Unlike prior work, the cross entropy part of our loss evaluates only seeds where labels are known while normalized cut softly evaluates consistency of all pixels. We focus on normalized cut loss where dense Gaussian kernel is efficiently implemented in linear time by fast Bilateral filtering. Our normalized cut loss approach to segmentation brings the quality of weakly-supervised training significantly closer to fully supervised methods.

Inspired by recent advances of deep learning in instance segmentation and object tracking, we introduce the concept of convnet-based guidance applied to video object segmentation. Our model proceeds on a per-frame basis, guided by the output of the previous frame towards the object of interest in the next frame. We demonstrate that highly accurate object segmentation in videos can be enabled by using a convolutional neural network (convnet) trained with static images only. The key component of our approach is a combination of offline and online learning strategies, where the former produces a refined mask from the previous frame estimate and the latter allows to capture the appearance of the specific object instance. Our method can handle different types of input annotations such as bounding boxes and segments while leveraging an arbitrary amount of annotated frames. Therefore our system is suitable for diverse applications with different requirements in terms of accuracy and efficiency. In our extensive evaluation, we obtain competitive results on three different datasets, independently from the type of input annotation.

In this work, we propose a novel approach to video segmentation that operates in bilateral space. We design a new energy on the vertices of a regularly sampled spatiotemporal bilateral grid, which can be solved efficiently using a standard graph cut label assignment. Using a bilateral formulation, the energy that we minimize implicitly approximates long-range, spatio-temporal connections between pixels while still containing only a small number of variables and only local graph edges. We compare to a number of recent methods, and show that our approach achieves state-of-the-art results on multiple benchmarks in a fraction of the runtime. Furthermore, our method scales linearly with image size, allowing for interactive feedback on real-world high resolution video.

Over the years, datasets and benchmarks have proven their fundamental importance in computer vision research, enabling targeted progress and objective comparisons in many fields. At the same time, legacy datasets may impend the evolution of a field due to saturated algorithm performance and the lack of contemporary, high quality data. In this work we present a new benchmark dataset and evaluation methodology for the area of video object segmentation. The dataset, named DAVIS (Densely Annotated VIdeo Segmentation), consists of fifty high quality, Full HD video sequences, spanning multiple occurrences of common video object segmentation challenges such as occlusions, motionblur and appearance changes. Each video is accompanied by densely annotated, pixel-accurate and per-frame ground truth segmentation. In addition, we provide a comprehensive analysis of several state-of-the-art segmentation approaches using three complementary metrics that measure the spatial extent of the segmentation, the accuracy of the silhouette contours and the temporal coherence. The results uncover strengths and weaknesses of current approaches, opening up promising directions for future works.

We present a novel approach to video segmentation using multiple object proposals. The problem is formulated as a minimization of a novel energy function defined over a fully connected graph of object proposals. Our model combines appearance with long-range point tracks, which is key to ensure robustness with respect to fast motion and occlusions over longer video sequences. As opposed to previous approaches based on object proposals, we do not seek the best per-frame object hypotheses to perform the segmentation. Instead, we combine multiple, potentially imperfect proposals to improve overall segmentation accuracy and ensure robustness to outliers. Overall, the basic algorithm consists of three steps. First, we generate a very large number of object proposals for each video frame using existing techniques. Next, we perform an SVM-based pruning step to retain only high quality proposals with sufficiently discriminative power. Finally, we determine the fore-and background classification by solving for the maximum a posteriori of a fully connected conditional random field, defined using our novel energy function. Experimental results on a well established dataset demonstrate that our method compares favorably to several recent state-of-the-art approaches.

Abstract We describe an algorithm for generating panoramic video from unstructured camera arrays. Artifact‐free panorama stitching is impeded by parallax between input views. Common strategies such as multi‐level blending or minimum energy seams produce seamless results on quasi‐static input. However, on video input these approaches introduce noticeable visual artifacts due to lack of global temporal and spatial coherence. In this paper we extend the basic concept of local warping for parallax removal. Firstly, we introduce an error measure with increased sensitivity to stitching artifacts in regions with pronounced structure. Using this measure, our method efficiently finds an optimal ordering of pair‐wise warps for robust stitching with minimal parallax artifacts. Weighted extrapolation of warps in non‐overlap regions ensures temporal stability, while at the same time avoiding visual discontinuities around transitions between views. Remaining global deformation introduced by the warps is spread over the entire panorama domain using constrained relaxation, while staying as close as possible to the original input views. In combination, these contributions form the first system for spatiotemporally stable panoramic video stitching from unstructured camera array input.

This paper extends Galerkin projection to a large class of non-polynomial functions typically encountered in graphics. We demonstrate the broad applicability of our approach by applying it to two strikingly different problems: fluid simulation and radiosity rendering, both using deforming meshes. Standard Galerkin projection cannot efficiently approximate these phenomena. Our approach, by contrast, enables the compact representation and approximation of these complex non-polynomial systems, including quotients and roots of polynomials. We rely on representing each function to be model-reduced as a composition of tensor products, matrix inversions, and matrix roots. Once a function has been represented in this form, it can be easily model-reduced, and its reduced form can be evaluated with time and memory costs dependent only on the dimension of the reduced space.

Saliency estimation has become a valuable tool in image processing. Yet, existing approaches exhibit considerable variation in methodology, and it is often difficult to attribute improvements in result quality to specific algorithm properties. In this paper we reconsider some of the design choices of previous methods and propose a conceptually clear and intuitive algorithm for contrast-based saliency estimation. Our algorithm consists of four basic steps. First, our method decomposes a given image into compact, perceptually homogeneous elements that abstract unnecessary detail. Based on this abstraction we compute two measures of contrast that rate the uniqueness and the spatial distribution of these elements. From the element contrast we then derive a saliency measure that produces a pixel-accurate saliency map which uniformly covers the objects of interest and consistently separates fore- and background. We show that the complete contrast and saliency estimation can be formulated in a unified way using high-dimensional Gaussian filters. This contributes to the conceptual simplicity of our method and lends itself to a highly efficient implementation with linear complexity. In a detailed experimental evaluation we analyze the contribution of each individual feature and show that our method outperforms all state-of-the-art approaches.