Search

We propose a novel system for portrait relighting and background replacement, which maintains high-frequency boundary details and accurately synthesizes the subject's appearance as lit by novel illumination, thereby producing realistic composite images for any desired scene. Our technique includes foreground estimation via alpha matting, relighting, and compositing. We demonstrate that each of these stages can be tackled in a sequential pipeline without the use of priors (e.g. known background or known illumination) and with no specialized acquisition techniques, using only a single RGB portrait image and a novel, target HDR lighting environment as inputs. We train our model using relit portraits of subjects captured in a light stage computational illumination system, which records multiple lighting conditions, high quality geometry, and accurate alpha mattes. To perform realistic relighting for compositing, we introduce a novel per-pixel lighting representation in a deep learning framework, which explicitly models the diffuse and the specular components of appearance, producing relit portraits with convincingly rendered non-Lambertian effects like specular highlights. Multiple experiments and comparisons show the effectiveness of the proposed approach when applied to in-the-wild images.

The light stage has been widely used in computer graphics for the past two decades, primarily to enable the relighting of human faces. By capturing the appearance of the human subject under different light sources, one obtains the light transport matrix of that subject, which enables image-based relighting in novel environments. However, due to the finite number of lights in the stage, the light transport matrix only represents a sparse sampling on the entire sphere. As a consequence, relighting the subject with a point light or a directional source that does not coincide exactly with one of the lights in the stage requires interpolation and resampling the images corresponding to nearby lights, and this leads to ghosting shadows, aliased specularities, and other artifacts. To ameliorate these artifacts and produce better results under arbitrary high-frequency lighting, this paper proposes a learning-based solution for the "super-resolution" of scans of human faces taken from a light stage. Given an arbitrary "query" light direction, our method aggregates the captured images corresponding to neighboring lights in the stage, and uses a neural network to synthesize a rendering of the face that appears to be illuminated by a "virtual" light source at the query location. This neural network must circumvent the inherent aliasing and regularity of the light stage data that was used for training, which we accomplish through the use of regularized traditional interpolation methods within our network. Our learned model is able to produce renderings for arbitrary light directions that exhibit realistic shadows and specular highlights, and is able to generalize across a wide variety of subjects., Comment: Siggraph Asia 2020

While the LED panels used in virtual production systems can display vibrant imagery with a wide color gamut, they produce problematic color shifts when used as lighting due to their peaky spectral output from narrow-band red, green, and blue LEDs. In this work, we present an improved color calibration process for virtual production stages which ameliorates this color rendition problem while also passing through accurate in-camera background colors. We do this by optimizing linear color correction transformations for 1) the LED panel pixels visible in the field of view of the camera, 2) the pixels outside the field of view of the camera illuminating the subjects, and, as a post-process, 3) the pixel values recorded by the camera. The result is that footage shot in an RGB LED panel virtual production stage can exhibit more accurate skin tones and costume colors while still reproducing the desired colors of the in-camera background., Comment: DigiPro 2022

We present a technique to reduce the dynamic range of an HDRI lighting environment map in an efficient, energy-preserving manner by spreading out the light of concentrated light sources. This allows us to display a reasonable approximation of the illumination of an HDRI map in a lighting reproduction system with limited dynamic range such as virtual production LED Stage. The technique identifies regions of the HDRI map above a given pixel threshold, dilates these regions until the average pixel value within each is below the threshold, and finally replaces each dilated region's pixels with the region's average pixel value. The new HDRI map contains the same energy as the original, spreads the light as little as possible, and avoids chromatic fringing.

We address the problem of recovering the shape and spatially-varying reflectance of an object from multi-view images (and their camera poses) of an object illuminated by one unknown lighting condition. This enables the rendering of novel views of the object under arbitrary environment lighting and editing of the object's material properties. The key to our approach, which we call Neural Radiance Factorization (NeRFactor), is to distill the volumetric geometry of a Neural Radiance Field (NeRF) [Mildenhall et al. 2020] representation of the object into a surface representation and then jointly refine the geometry while solving for the spatially-varying reflectance and environment lighting. Specifically, NeRFactor recovers 3D neural fields of surface normals, light visibility, albedo, and Bidirectional Reflectance Distribution Functions (BRDFs) without any supervision, using only a re-rendering loss, simple smoothness priors, and a data-driven BRDF prior learned from real-world BRDF measurements. By explicitly modeling light visibility, NeRFactor is able to separate shadows from albedo and synthesize realistic soft or hard shadows under arbitrary lighting conditions. NeRFactor is able to recover convincing 3D models for free-viewpoint relighting in this challenging and underconstrained capture setup for both synthetic and real scenes. Qualitative and quantitative experiments show that NeRFactor outperforms classic and deep learning-based state of the art across various tasks. Our videos, code, and data are available at people.csail.mit.edu/xiuming/projects/nerfactor/., Camera-ready version for SIGGRAPH Asia 2021. Project Page: https://people.csail.mit.edu/xiuming/projects/nerfactor/

We present "The Relightables", a volumetric capture system for photorealistic and high quality relightable full-body performance capture. While significant progress has been made on volumetric capture systems, focusing on 3D geometric reconstruction with high resolution textures, much less work has been done to recover photometric properties needed for relighting. Results from such systems lack high-frequency details and the subject's shading is prebaked into the texture. In contrast, a large body of work has addressed relightable acquisition for image-based approaches, which photograph the subject under a set of basis lighting conditions and recombine the images to show the subject as they would appear in a target lighting environment. However, to date, these approaches have not been adapted for use in the context of a high-resolution volumetric capture system. Our method combines this ability to realistically relight humans for arbitrary environments, with the benefits of free-viewpoint volumetric capture and new levels of geometric accuracy for dynamic performances. Our subjects are recorded inside a custom geodesic sphere outfitted with 331 custom color LED lights, an array of high-resolution cameras, and a set of custom high-resolution depth sensors. Our system innovates in multiple areas: First, we designed a novel active depth sensor to capture 12.4 MP depth maps, which we describe in detail. Second, we show how to design a hybrid geometric and machine learning reconstruction pipeline to process the high resolution input and output a volumetric video. Third, we generate temporally consistent reflectance maps for dynamic performers by leveraging the information contained in two alternating color gradient illumination images acquired at 60Hz. Multiple experiments, comparisons, and applications show that The Relightables significantly improves upon the level of realism in placing volumetrically captured human performances into arbitrary CG scenes.

We present a novel technique to relight images of human faces by learning a model of facial reflectance from a database of 4D reflectance field data of several subjects in a variety of expressions and viewpoints. Using our learned model, a face can be relit in arbitrary illumination environments using only two original images recorded under spherical color gradient illumination. The output of our deep network indicates that the color gradient images contain the information needed to estimate the full 4D reflectance field, including specular reflections and high frequency details. While capturing spherical color gradient illumination still requires a special lighting setup, reduction to just two illumination conditions allows the technique to be applied to dynamic facial performance capture. We show side-by-side comparisons which demonstrate that the proposed system outperforms the state-of-the-art techniques in both realism and speed.

We present a system for acquiring, processing, and rendering panoramic light field still photography for display in Virtual Reality (VR). We acquire spherical light field datasets with two novel light field camera rigs designed for portable and efficient light field acquisition. We introduce a novel real-time light field reconstruction algorithm that uses a per-view geometry and a disk-based blending field. We also demonstrate how to use a light field prefiltering operation to project from a high-quality offline reconstruction model into our real-time model while suppressing artifacts. We introduce a practical approach for compressing light fields by modifying the VP9 video codec to provide high quality compression with real-time, random access decompression. We combine these components into a complete light field system offering convenient acquisition, compact file size, and high-quality rendering while generating stereo views at 90Hz on commodity VR hardware. Using our system, we built a freely available light field experience application called Welcome to Light Fields featuring a library of panoramic light field stills for consumer VR which has been downloaded over 15,000 times., Comment: 15 pages, 14 figures, 2 tables, accepted by SIGGRAPH Asia 2018, low-resolution version

Out-of-focus points of light, when obscured by out-of-focus occluders, become “eclipsed” by a sharply-focused traveling occlusion edge which can move at a speed different from that of the occluder, and even in the opposite direction. The phenomenon can produce interesting visual effects in photos and motion pictures.

This Immersive Pavilion installation introduces our new system for capturing, reconstructing, compressing, and rendering light field video content. By leveraging DeepView, a recently introduced view synthesis algorithm, our system can reconstruct challenging scenes with view-dependent reflections, semi-transparent surfaces, and near-field objects as close as 34 cm to the surface of our 46 camera capture rig. Improving upon past light field video systems that required specialized storage and graphics hardware for playback, our compressed videos can be rendered in a web browser or on mobile VR headsets while being streamed over a gigabit network connection. This makes ours the first system to encode high quality light field video at sufficiently low bandwidth for internet streaming.

We present a system for capturing, reconstructing, compressing, and rendering high quality immersive light field video. We accomplish this by leveraging the recently introduced DeepView view interpolation algorithm, replacing its underlying multi-plane image (MPI) scene representation with a collection of spherical shells that are better suited for representing panoramic light field content. We further process this data to reduce the large number of shell layers to a small, fixed number of RGBA+depth layers without significant loss in visual quality. The resulting RGB, alpha, and depth channels in these layers are then compressed using conventional texture atlasing and video compression techniques. The final compressed representation is lightweight and can be rendered on mobile VR/AR platforms or in a web browser. We demonstrate light field video results using data from the 16-camera rig of [Pozo et al. 2019] as well as a new low-cost hemispherical array made from 46 synchronized action sports cameras. From this data we produce 6 degree of freedom volumetric videos with a wide 70 cm viewing baseline, 10 pixels per degree angular resolution, and a wide field of view, at 30 frames per second video frame rates. Advancing over previous work, we show that our system is able to reproduce challenging content such as view-dependent reflections, semi-transparent surfaces, and near-field objects as close as 34 cm to the surface of the camera rig.

The light transport (LT) of a scene describes how it appears under different lighting and viewing directions, and complete knowledge of a scene's LT enables the synthesis of novel views under arbitrary lighting. In this paper, we focus on image-based LT acquisition, primarily for human bodies within a light stage setup. We propose a semi-parametric approach to learn a neural representation of LT that is embedded in the space of a texture atlas of known geometric properties, and model all non-diffuse and global LT as residuals added to a physically-accurate diffuse base rendering. In particular, we show how to fuse previously seen observations of illuminants and views to synthesize a new image of the same scene under a desired lighting condition from a chosen viewpoint. This strategy allows the network to learn complex material effects (such as subsurface scattering) and global illumination, while guaranteeing the physical correctness of the diffuse LT (such as hard shadows). With this learned LT, one can relight the scene photorealistically with a directional light or an HDRI map, synthesize novel views with view-dependent effects, or do both simultaneously, all in a unified framework using a set of sparse, previously seen observations. Qualitative and quantitative experiments demonstrate that our neural LT (NLT) outperforms state-of-the-art solutions for relighting and view synthesis, without separate treatment for both problems that prior work requires., Camera-ready version for TOG 2021. Project Page: http://nlt.csail.mit.edu/

The increasing demand for 3D content in augmented and virtual reality has motivated the development of volumetric performance capture systemsnsuch as the Light Stage. Recent advances are pushing free viewpoint relightable videos of dynamic human performances closer to photorealistic quality. However, despite significant efforts, these sophisticated systems are limited by reconstruction and rendering algorithms which do not fully model complex 3D structures and higher order light transport effects such as global illumination and sub-surface scattering. In this paper, we propose a system that combines traditional geometric pipelines with a neural rendering scheme to generate photorealistic renderings of dynamic performances under desired viewpoint and lighting. Our system leverages deep neural networks that model the classical rendering process to learn implicit features that represent the view-dependent appearance of the subject independent of the geometry layout, allowing for generalization to unseen subject poses and even novel subject identity. Detailed experiments and comparisons demonstrate the efficacy and versatility of our method to generate high-quality results, significantly outperforming the existing state-of-the-art solutions.

We present a multi-view stereo reconstruction technique that directly produces a complete high-fidelity head model with consistent facial mesh topology. While existing techniques decouple shape estimation and facial tracking, our framework jointly optimizes for stereo constraints and consistent mesh parameterization. Our method is therefore free from drift and fully parallelizable for dynamic facial performance capture. We produce highly detailed facial geometries with artist-quality UV parameterization, including secondary elements such as eyeballs, mouth pockets, nostrils, and the back of the head. Our approach consists of deforming a common template model to match multi-view input images of the subject, while satisfying cross-view, cross-subject, and cross-pose consistencies using a combination of 2D landmark detection, optical flow, and surface and volumetric Laplacian regularization. Since the flow is never computed between frames, our method is trivially parallelized by processing each frame independently. Accurate rigid head pose is extracted using a PCA-based dimension reduction and denoising scheme. We demonstrate high-fidelity performance capture results with challenging head motion and complex facial expressions around eye and mouth regions. While the quality of our results is on par with the current state-of-the-art, our approach can be fully parallelized, does not suffer from drift, and produces face models with production-quality mesh topologies.

We present a portable multi-camera system for recording panoramic light field video content. The proposed system captures wide baseline (0.8 meters), high resolution (>15 pixels per degree), large field of view (>220°) light fields at 30 frames per second. The array contains 47 time-synchronized cameras distributed on the surface of a hemispherical, 0.92 meter diameter plastic dome. We use commercially available action sports cameras (Yi 4k) mounted inside the dome using 3D printed brackets. The dome, mounts, triggering hardware and cameras are inexpensive and the array itself is easy to fabricate. Using modern view interpolation algorithms we can render objects as close as 33-cm to the surface of the array.

Motivated by recent availability of augmented and virtual reality platforms, we tackle the challenging problem of immersive storytelling experiences on mobile devices. In particular, we show an end-to-end system to generate 3D assets that enable real-time rendering of an opera on high end mobile phones. We call our system AR-ia and in this paper we walk through the main components and technical challenges of such a system, showing how to deliver an immersive mixed reality experience in every user’s living room.

We present a learning-based method to infer plausible high dynamic range (HDR), omnidirectional illumination given an unconstrained, low dynamic range (LDR) image from a mobile phone camera with a limited field of view (FOV). For training data, we collect videos of various reflective spheres placed within the camera's FOV, leaving most of the background unoccluded, leveraging that materials with diverse reflectance functions reveal different lighting cues in a single exposure. We train a deep neural network to regress from the LDR background image to HDR lighting by matching the LDR ground truth sphere images to those rendered with the predicted illumination using image-based relighting, which is differentiable. Our inference runs at interactive frame rates on a mobile device, enabling realistic rendering of virtual objects into real scenes for mobile mixed reality. Training on automatically exposed and white-balanced videos, we improve the realism of rendered objects compared to the state-of-the art methods for both indoor and outdoor scenes.

We present a variety of new compositing techniques using Multi-plane Images (MPI's) [Zhou et al. 2018] derived from footage shot with an inexpensive and portable light field video camera array. The effects include camera stabilization, foreground object removal, synthetic depth of field, and deep compositing. Traditional compositing is based around layering RGBA images to visually integrate elements into the same scene, and often requires manual 2D and/or 3D artist intervention to achieve realism in the presence of volumetric effects such as smoke or splashing water. We leverage the newly introduced DeepView solver [Flynn et al. 2019] and a light field camera array to generate MPIs stored in the DeepEXR format for compositing with realistic spatial integration and a simple workflow which offers new creative capabilities. We demonstrate using this technique by combining footage that would otherwise be very challenging and time intensive to achieve when using traditional techniques, with minimal artist intervention.

We present a novel approach to view synthesis using multiplane images (MPIs). Building on recent advances in learned gradient descent, our algorithm generates an MPI from a set of sparse camera viewpoints. The resulting method incorporates occlusion reasoning, improving performance on challenging scene features such as object boundaries, lighting reflections, thin structures, and scenes with high depth complexity. We show that our method achieves high-quality, state-of-the-art results on two datasets: the Kalantari light field dataset, and a new camera array dataset, Spaces, which we make publicly available., See https://augmentedperception.github.io/deepview/ for more results, video and an interactive viewer

Lighting plays a central role in conveying the essence and depth of the subject in a portrait photograph. Professional photographers will carefully control the lighting in their studio to manipulate the appearance of their subject, while consumer photographers are usually constrained to the illumination of their environment. Though prior works have explored techniques for relighting an image, their utility is usually limited due to requirements of specialized hardware, multiple images of the subject under controlled or known illuminations, or accurate models of geometry and reflectance. To this end, we present a system for portrait relighting: a neural network that takes as input a single RGB image of a portrait taken with a standard cellphone camera in an unconstrained environment, and from that image produces a relit image of that subject as though it were illuminated according to any provided environment map. Our method is trained on a small database of 18 individuals captured under different directional light sources in a controlled light stage setup consisting of a densely sampled sphere of lights. Our proposed technique produces quantitatively superior results on our dataset's validation set compared to prior works, and produces convincing qualitative relighting results on a dataset of hundreds of real-world cellphone portraits. Because our technique can produce a 640 $\times$ 640 image in only 160 milliseconds, it may enable interactive user-facing photographic applications in the future., Comment: SIGGRAPH 2019 Technical Paper accepted

We present a practical framework for reproducing omnidirectional incident illumination conditions with complex spectra using a light stage with multispectral LED lights. For lighting acquisition, we augment standard RGB panoramic photography with one or more observations of a color chart with numerous reflectance spectra. We then solve for how to drive the multispectral light sources so that they best reproduce the appearance of the color charts in the original lighting. Even when solving for non-negative intensities, we show that accurate lighting reproduction is achievable using just four or six distinct LED spectra for a wide range of incident illumination spectra. A significant benefit of our approach is that it does not require the use of specialized equipment (other than the light stage) such as monochromators, spectroradiometers, or explicit knowledge of the LED power spectra, camera spectral response functions, or color chart reflectance spectra. We describe two simple devices for multispectral lighting capture, one for slow measurements of detailed angular spectral detail, and one for fast measurements with coarse angular detail. We validate the approach by realistically compositing real subjects into acquired lighting environments, showing accurate matches to how the subject would actually look within the environments, even for those including complex multispectral illumination. We also demonstrate dynamic lighting capture and playback using the technique.

We propose a variant to polarized gradient illumination facial scanning which uses monochrome instead of color cameras to achieve more efficient and higher-resolution results. In typical polarized gradient facial scanning, sub-millimeter geometric detail is acquired by photographing the subject in eight or more polarized spherical gradient lighting conditions made with white LEDs, and RGB cameras are used to acquire color texture maps of the subject's appearance. In our approach, we replace the color cameras and white LEDs with monochrome cameras and multispectral, colored LEDs, leveraging that color images can be formed from successive monochrome images recorded under different illumination colors. While a naive extension of the scanning process to this setup would require multiplying the number of images by number of color channels, we show that the surface detail maps can be estimated directly from monochrome imagery, so that only an additional n photographs are required, where n is the number of added spectral channels. We also introduce a new multispectral optical flow approach to align images across spectral channels in the presence of slight subject motion. Lastly, for the case where a capture system's white light sources are polarized and its multispectral colored LEDs are not, we introduce the technique of multispectral polarization promotion, where we estimate the cross- and parallel-polarized monochrome images for each spectral channel from their corresponding images under a full sphere of even, unpolarized illumination. We demonstrate that this technique allows us to efficiently acquire a full color (or even multispectral) facial scan using monochrome cameras, unpolarized multispectral colored LEDs, and polarized white LEDs.

Light fields can provide transportive immersive experiences with a level of realism unsurpassed by any other imaging technology. Within a limited viewing volume, light fields accurately reproduce stereo parallax, motion parallax, reflections, refractions, and volumetric effects for real-world scenes. While light fields have been explored in computer graphics since the mid-90's [Gortier et al. 1996; Levoy and Hanrahan 1996], practical systems for recording, processing, and delivering high quality light field experiences have remained out of reach.In this talk, we describe the hardware and software developed to produce the Welcome to Light Fields virtual reality experience, the first immersive light field experience that can be downloaded and viewed on VR-ready computers with HTC Vive, Oculus Rift, and Windows Mixed Reality HMDs. This piece enables the user to step into a collection of panoramic light field still photographs and be guided through a number of light field environments, including the flight deck of Space Shuttle Discovery. Welcome to Light Fields is a free VR app. on the Steam Store that was launched in March 2018 [Google VR 2018].This piece is the culmination of a multi-year effort to develop consumer-ready light field technology. We built new camera rigs for light field capture and new algorithms for light field rendering. In this talk, we delve into the details from capture to rendering and the steps in between.We emphasize that our system generates stereoscopic views of a light field for use in current generation consumer VR HMDs. As future work, it could be of interest to display such datasets on near eye light field displays [Lanman and Luebke 2013] to render effects of visual accommodation.

Light Fields let us experience freedom of motion and realistic reflections and translucence like never before in VR. Explore the Gamble House, Mosaic Tile House, and Space Shuttle Discovery. These navigable light field stills showcase the emerging technology Google is using to power its next generation of VR content.

We present a learning-based approach for synthesizing facial geometry at medium and fine scales from diffusely-lit facial texture maps. When applied to an image sequence, the synthesized detail is temporally coherent. Unlike current state-of-the-art methods [17, 5], which assume "dark is deep", our model is trained with measured facial detail collected using polarized gradient illumination in a Light Stage [20]. This enables us to produce plausible facial detail across the entire face, including where previous approaches may incorrectly interpret dark features as concavities such as at moles, hair stubble, and occluded pores. Instead of directly inferring 3D geometry, we propose to encode fine details in high-resolution displacement maps which are learned through a hybrid network adopting the state-of-the-art image-to-image translation network [29] and super resolution network [43]. To effectively capture geometric detail at both mid- and high frequencies, we factorize the learning into two separate sub-networks, enabling the full range of facial detail to be modeled. Results from our learning-based approach compare favorably with a high-quality active facial scanhening technique, and require only a single passive lighting condition without a complex scanning setup.

We present a technique for synthesizing the effects of skin microstructure deformation by anisotropically convolving a high-resolution displacement map to match normal distribution changes in measured skin samples. We use a 10-micron resolution scanning technique to measure several in vivo skin samples as they are stretched and compressed in different directions, quantifying how stretching smooths the skin and compression makes it rougher. We tabulate the resulting surface normal distributions, and show that convolving a neutral skin microstructure displacement map with blurring and sharpening filters can mimic normal distribution changes and microstructure deformations. We implement the spatially-varying displacement map filtering on the GPU to interactively render the effects of dynamic microgeometry on animated faces obtained from high-resolution facial scans.

We present a process for rendering a realistic facial performance with control of viewpoint and illumination. The performance is based on one or more high-quality geometry and reflectance scans of an actor in static poses, driven by one or more video streams of a performance. We compute optical flow correspondences between neighboring video frames, and a sparse set of correspondences between static scans and video frames. The latter are made possible by leveraging the relightability of the static 3D scans to match the viewpoint(s) and appearance of the actor in videos taken in arbitrary environments. As optical flow tends to compute proper correspondence for some areas but not others, we also compute a smoothed, per-pixel confidence map for every computed flow, based on normalized cross-correlation. These flows and their confidences yield a set of weighted triangulation constraints among the static poses and the frames of a performance. Given a single artist-prepared face mesh for one static pose, we optimally combine the weighted triangulation constraints, along with a shape regularization term, into a consistent 3D geometry solution over the entire performance that is drift free by construction. In contrast to previous work, even partial correspondences contribute to drift minimization, for example, where a successful match is found in the eye region but not the mouth. Our shape regularization employs a differential shape term based on a spatially varying blend of the differential shapes of the static poses and neighboring dynamic poses, weighted by the associated flow confidences. These weights also permit dynamic reflectance maps to be produced for the performance by blending the static scan maps. Finally, as the geometry and maps are represented on a consistent artist-friendly mesh, we render the resulting high-quality animated face geometry and animated reflectance maps using standard rendering tools.

We present a technique for modeling the vellus hair over the face based on observations of asperity scattering along a subject's silhouette. We photograph the backlit subject in profile and three-quarters views with a high-resolution DSLR camera to observe the vellus hair on the side and front of the face and separately acquire a 3D scan of the face geometry and texture. We render a library of backlit vellus hair patch samples with different geometric parameters such as density, orientation, and curvature, and we compute image statistics for each set of parameters. We trace the silhouette contour in each face image and straighten the backlit hair silhouettes using image resampling. We compute image statistics for each section of the facial silhouette and determine which set of hair modeling parameters best matches the statistics. We then generate a complete set of vellus hairs for the face by interpolating and extrapolating the matched parameters over the skin. We add the modeled vellus hairs to the 3D facial scan and generate renderings under novel lighting conditions, generally matching the appearance of real photographs.

We present a technique for improving the alpha matting of challenging green-screen video sequences involving hair strands. As hair strands are thin and can be semi-translucent, they are especially hard to separate from a background. However, they appear as extended lines and thus have a strong response when convolved with oriented filters, even in the presence of noise. We leverage this oriented filter response to robustly locate hair strands within each frame of an actor's performance filmed in front of a green-screen. We demonstrate using production video footage that individual hair fibers excluded from a coarse artist's matte can be located and then added to the foreground element, qualitatively improving the composite result without added manual labor.

Computer vision is a rapidly evolving discipline. It includes methods for acquiring, processing, and understanding still images and video to model, replicate, and sometimes, exceed human vision and perform useful tasks.Computer vision will be commonly used for a broad range of services in upcoming devices, and implemented in everything from movies, smartphones, cameras, drones and more. Demand for CV is driving the evolution of image sensors, mobile processors, operating systems, application software, and device form factors in order to meet the needs of upcoming applications and services that benefit from computer vision. The resulting impetus means rapid advancements in:• visual computing performance• object recognition effectiveness• speed and responsiveness• power efficiency• video image quality improvement• real-time 3D reconstruction• pre-scanning for movie animation• image stabilization• immersive experiences• and more...Comprised of innovation leaders of computer vision, this panel will cover recent developments, as well as how CV will be enabled and used in 2016 and beyond.

We demonstrate the sufficiency of using as few as five LEDs of distinct spectra for multispectral lighting reproduction and solve for the optimal set of five from 11 such commercially available LEDs. We leverage published spectral reflectance, illuminant, and camera spectral sensitivity datasets to show that two approaches of lighting reproduction, matching illuminant spectra directly and matching material color appearance observed by one or more cameras or a human observer, yield the same LED selections. Our proposed optimal set of five LEDs includes red, green, and blue with narrow emission spectra, along with white and amber with broader spectra.

We present a system for creating and displaying interactive life-sized 3D digital humans based on pre-recorded interviews. We use 30 cameras and an extensive list of questions to record a large set of video responses. Users access videos through a natural conversation interface that mimics face-to-face interaction. Recordings of answers, listening and idle behaviors are linked together to create a persistent visual image of the person throughout the interaction. The interview subjects are rendered using flowed light fields and shown life-size on a special rear-projection screen with an array of 216 video projectors. The display allows multiple users to see different 3D perspectives of the subject in proper relation to their viewpoints, without the need for stereo glasses. The display is effective for interactive conversations since it provides 3D cues such as eye gaze and spatial hand gestures.

Creating and animating realistic 3D human faces is an important element of virtual reality, video games, and other areas that involve interactive 3D graphics. In this paper, we propose a system to generate photorealistic 3D blendshape-based face models automatically using only a single consumer RGB-D sensor. The capture and processing requires no artistic expertise to operate, takes 15 seconds to capture and generate a single facial expression, and approximately 1 minute of processing time per expression to transform it into a blendshape model. Our main contributions include a complete end-to-end pipeline for capturing and generating photorealistic blendshape models automatically and a registration method that solves dense correspondences between two face scans by utilizing facial landmarks detection and optical flows. We demonstrate the effectiveness of the proposed method by capturing different human subjects with a variety of sensors and puppeteering their 3D faces with real-time facial performance retargeting. The rapid nature of our method allows for just-in-time construction of a digital face. To that end, we also integrated our pipeline with a virtual reality facial performance capture system that allows dynamic embodiment of the generated faces despite partial occlusion of the user's real face by the head-mounted display.

We present a near-instant method for acquiring facial geometry and reflectance using a set of commodity DSLR cameras and flashes. Our setup consists of twenty-four cameras and six flashes which are fired in rapid succession with subsets of the cameras. Each camera records only a single photograph and the total capture time is less than the 67ms blink reflex. The cameras and flashes are specially arranged to produce an even distribution of specular highlights on the face. We employ this set of acquired images to estimate diffuse color, specular intensity, specular exponent, and surface orientation at each point on the face. We further refine the facial base geometry obtained from multi-view stereo using estimated diffuse and specular photometric information. This allows final submillimeter surface mesostructure detail to be obtained via shape-from-specularity. The final system uses commodity components and produces models suitable for authoring high-quality digital human characters.

One of the main challenges in facial expression recognition is illumination invariance. Our long-term goal is to develop a system for automatic facial expression recognition that is robust to light variations. In this paper, we introduce a novel 3D Relightable Facial Expression (ICT-3DRFE) database that enables experimentation in the fields of both computer graphics and computer vision. The database contains 3D models for 23 subjects and 15 expressions, as well as photometric information that allow for photorealistic rendering. It is also facial action units annotated, using FACS standards. Using the ICT-3DRFE database we create an image set of different expressions/illuminations to study the effect of illumination on automatic expression recognition. We compared the output scores from automatic recognition with expert FACS annotations and found that they agree when the illumination is uniform. Our results show that the output distribution of the automatic recognition can change significantly with light variations and sometimes causes the discrimination of two different expressions to be diminished. We propose a ratio-based light transfer method, to factor out unwanted illuminations from given images and show that it reduces the effect of illumination on expression recognition.

We present a system for recording a live dynamic facial performance, capturing highly detailed geometry and spatially varying diffuse and specular reflectance information for each frame of the performance. The result is a reproduction of the performance that can be rendered from novel viewpoints and novel lighting conditions, achieving photorealistic integration into any virtual environment. Dynamic performances are captured directly, without the need for any template geometry or static geometry scans, and processing is completely automatic, requiring no human input or guidance. Our key contributions are a heuristic for estimating facial reflectance information from gradient illumination photographs, and a geometry optimization framework that maximizes a principled likelihood function combining multi-view stereo correspondence and photometric stereo, using multi-resolution belief propagation. The output of our system is a sequence of geometries and reflectance maps, suitable for rendering in off-the-shelf software. We show results from our system rendered under novel viewpoints and lighting conditions, and validate our results by demonstrating a close match to ground truth photographs.

We present a facial motion retargeting method that enables the control of a blendshape rig according to marker-based motion capture data. The main purpose of the proposed technique is to allow a blendshape rig to create facial expressions, which conforms best to the current motion capture input, regardless the underlying blendshape poses. In other words, even though all of the blendshape poses may comprise symmetrical facial expressions only, our method is still able to create asymmetrical expressions without physically splitting any of them into more local blendshape poses. An automatic segmentation technique based on the analysis of facial motion is introduced to create facial regions for local retargeting. We also show that it is possible to blend normal maps for rendering in the same framework. Rendering with the blended normal map significantly improves surface appearance and details. Copyright © 2011 John Wiley & Sons, Ltd.

We present a novel technique for acquiring detailed facial geometry of a dynamic performance using extended spherical gradient illumination. Key to our method is a new algorithm for jointly aligning two photographs, under a gradient illumination condition and its complement, to a full-on tracking frame, providing dense temporal correspondences under changing lighting conditions. We employ a two-step algorithm to reconstruct detailed geometry for every captured frame. In the first step, we coalesce information from the gradient illumination frames to the full-on tracking frame, and form a temporally aligned photometric normal map, which is subsequently combined with dense stereo correspondences yielding a detailed geometry. In a second step, we propagate the detailed geometry back to every captured instance guided by the previously computed dense correspondences. We demonstrate reconstructed dynamic facial geometry, captured using moderate to video rates of acquisition, for every captured frame.

We describe a system for high-resolution capture of moving 3D geometry, beginning with dynamic normal maps from multiple views. The normal maps are captured using active shape-from-shading (photometric stereo), with a large lighting dome providing a series of novel spherical lighting configurations. To compensate for low-frequency deformation, we perform multi-view matching and thin-plate spline deformation on the initial surfaces obtained by integrating the normal maps. Next, the corrected meshes are merged into a single mesh using a volumetric method. The final output is a set of meshes, which were impossible to produce with previous methods. The meshes exhibit details on the order of a few millimeters, and represent the performance over human-size working volumes at a temporal resolution of 60Hz.

This paper presents a novel method for estimating specular roughness and tangent vectors, per surface point, from polarized second order spherical gradient illumination patterns. We demonstrate that for isotropic BRDFs, only three second order spherical gradients are sufficient to robustly estimate spatially varying specular roughness. For anisotropic BRDFs, an additional two measurements yield specular roughness and tangent vectors per surface point. We verify our approach with different illumination configurations which project both discrete and continuous fields of gradient illumination. Our technique provides a direct estimate of the per-pixel specular roughness and thus does not require off-line numerical optimization that is typical for the measure-and-fit approach to classical BRDF modeling.

In this article we propose a new framework for capturing light transport data of a real scene, based on the recently developed theory of compressive sensing. Compressive sensing offers a solid mathematical framework to infer a sparse signal from a limited number of nonadaptive measurements. Besides introducing compressive sensing for fast acquisition of light transport to computer graphics, we develop several innovations that address specific challenges for image-based relighting, and which may have broader implications. We develop a novel hierarchical decoding algorithm that improves reconstruction quality by exploiting interpixel coherency relations. Additionally, we design new nonadaptive illumination patterns that minimize measurement noise and further improve reconstruction quality. We illustrate our framework by capturing detailed high-resolution reflectance fields for image-based relighting.

High-end facial performance capture solutions typically use head-mounted camera systems which provide one or more close-up video streams of each actor's performance. These provide clear views of each actor's performance, but can be bulky, uncomfortable, get in the way of sight lines, and prevent actors from getting close to each other. To address this, we propose a virtual head-mounted camera system: an array of cameras placed around around the performance capture volume which automatically track zoomed-in, sharply focussed, high-resolution views of the each actor's face from a multitude of directions. The resulting imagery can be used in conjunction with body motion capture data to derive nuanced facial performances without head-mounted cameras.

Todays most compelling virtual reality experiences shift the users viewpoint within the virtual environment based on input from a head-tracking system, giving a compelling sense of motion parallax. While this is straightforward for computer generated scenes, photographic VR content generally does not provide motion parallax in response to head motion. Even 360° stereo panoramas, which offer separated left and right views, fail to allow the vantage point to change in response to head motion.

Creating and animating a realistic 3D human face is an important task in computer graphics. The capability of capturing the 3D face of a human subject and reanimate it quickly will find many applications in games, training simulations, and interactive 3D graphics. We demonstrate a system to capture photorealistic 3D faces and generate the blendshape models automatically using only a single commodity RGB-D sensor. Our method can rapidly generate a set of expressive facial poses from a single depth sensor, such as a Microsoft Kinect version 1, and requires no artistic expertise in order to process those scans. The system takes only a matter of seconds to capture and produce a 3D facial pose and only requires a few minutes of processing time to transform it into a blendshape-compatible model. Our main contributions include an end-to-end pipeline for capturing and generating face blendshape models automatically, and a registration method that solves dense correspondences between two face scans by utilizing facial landmarks detection and optical flows. We demonstrate the effectiveness of the proposed method by capturing different human subjects and puppeteering their 3D faces in an animation system with real-time facial performance retargeting.

On June 9th, 2014, we traveled to the State Dining Room of The White House to create a 3D Portrait of President Barack Obama using state-of-the-art 3D scanning and printing technology, producing the modern equivalent of the plaster life masks of President Lincoln from the 1860's.