We present an implicit video representation for occlusions, appearance, and motion disentanglement from monocular videos, which we call Video SPatiotemporal Splines (VideoSPatS). Unlike previous methods that map time and coordinates to deformation and canonical colors, our VideoSPatS maps input coordinates into Spatial and Color Spline deformation fields ${\mathcal{D}_{\text{s}}}$ and ${\mathcal{D}_c}$, which disentangle motion and appearance in videos. With spline-based parametrization, our method naturally generates temporally consistent flow and guarantees long-term temporal consistency, which is crucial for convincing video editing. Using multiple prediction branches, our VideoSPatS model also performs layer separation between the latent video and the selected occluder. By disentangling occlusions, appearance, and motion, our method enables better spatiotemporal modeling and editing of diverse videos, including in-the-wild talking head videos with challenging occlusions, shadows, and specularities while maintaining an appropriate canonical space for editing. We also present general video modeling results on the DAVIS and CoDeF datasets, as well as our own talking head video dataset collected from open-source web videos. Extensive ablations show the combination of ${\mathcal{D}_{\text{s}}}$ and ${\mathcal{D}_c}$ under neural splines can overcome motion and appearance ambiguities, paving the way for more advanced video editing models. Visit our project site1.

Co-speech gesture generation is crucial for creating lifelike avatars and enhancing human-computer interactions by synchronizing gestures with speech. Despite recent advancements, existing methods struggle with accurately identifying the rhythmic or semantic triggers from audio for generating contextualized gesture patterns and achieving pixel-level realism. To address these challenges, we introduce Contextual Gesture, a framework that improves co-speech gesture video generation through three innovative components: (1) a chronological speech-gesture alignment that temporally connects two modalities, (2) a contextualized gesture tokenization that incorporate speech context into motion pattern representation through distillation, and (3) a structure-aware refinement module that employs edge connection to link gesture keypoints to improve video generation. Our extensive experiments demonstrate that Contextual Gesture not only produces realistic and speech-aligned gesture videos but also supports long-sequence generation and video gesture editing applications, shown in Fig.1

The two popular datasets ScanRefer [20] and ReferIt3D [5] connect natural language to real-world 3D scenes. In this paper, we curate a complementary dataset extending both the aforementioned ones. We associate all objects mentioned in a referential sentence with their underlying instances inside a 3D scene. In contrast, previous work did this only for a single object per sentence. Our Scan Entities in 3D (ScanEnts3D) dataset provides explicit correspondences between 369k objects across 84k referential sentences, covering 705 real-world scenes. We propose novel architecture modifications and losses that enable learning from this new type of data and improve the performance for both neural listening and language generation. For neural listening, we improve the SoTA in both the Nr3D and ScanRefer benchmarks by 4.3% and 5.0%, respectively. For language generation, we improve the SoTA by 13.2 CIDEr points on the Nr3D benchmark. For both of these tasks, the new type of data is only used to improve training, but no additional annotations are required at inference time. Our introduced dataset is available on the project’s webpage at https://scanents3d.github.io/.

We propose a novel approach for unsupervised 3D animation of non-rigid deformable objects. Our method learns the 3D structure and dynamics of objects solely from single-view RGB videos, and can decompose them into semantically meaningful parts that can be tracked and animated. Using a 3D autodecoder framework, paired with a keypoint estimator via a differentiable PnP algorithm, our model learns the underlying object geometry and parts decomposition in an entirely unsupervised manner. This allows it to perform 3D segmentation, 3D keypoint estimation, novel view synthesis, and animation. We primarily evaluate the framework on two video datasets: VoxCeleb 2562and TEDXPeople 2562. In addition, on the Cats 2562image dataset, we show it even learns compelling 3D geometry from still images. Finally, we show our model can obtain animatable 3D objects from a single or few images11Code and visual results available on our project website: https://snap-research.github.io/unsupervised-volumetric-animation..

Diffusion-based methods have shown impressive visual results in the text-to-image domain. They first learn a latent space using an autoencoder, then run a denoising process on the bottleneck to generate new samples. However, learning an autoencoder requires substantial data in the target domain. Such data is scarce for 3D generation, prohibiting the learning of large-scale diffusion models for 3D synthesis. We present a novel approach to the generation of static and articulated 3D assets that has a 3D autodecoder at its core. The 3D autodecoder framework embeds properties learned from the target dataset in the latent space, which can then be decoded into a volumetric representation for rendering view-consistent appearance and geometry. We then identify the appropriate intermediate volumetric latent space, and introduce robust normalization and de-normalization operations to learn a 3D diffusion from 2D images or monocular videos of rigid or articulated objects. Our approach is flexible enough to use either existing camera supervision or no camera information at all -- instead efficiently learning it during training. Our evaluations demonstrate that our generation results outperform state-of-the-art alternatives on various benchmark datasets and metrics, including multi-view image datasets of synthetic objects, real in-the-wild videos of moving people, and a large-scale, real video dataset of static objects.

We present Control-NeRF1, a method for performing flexible, 3D-aware image content manipulation while enabling high-quality novel view synthesis, from a set of posed input images. NeRF-based approaches [23] are effective for novel view synthesis, however such models memorize the radiance for every point in a scene within a neural network. Since these models are scene-specific and lack a 3D scene representation, classical editing such as shape manipulation, or combining scenes is not possible. While there are some recent hybrid approaches that combine NeRF with external scene representations such as sparse voxels, planes, hash tables, etc. [16], [5], [24], [9], they focus mostly on efficiency and don't explore the scene editing and manipulation capabilities of hybrid approaches. With the aim of exploring controllable scene representations for novel view synthesis, our model couples learnt scene-specific 3D feature volumes with a general NeRF rendering network. We can generalize to novel scenes by optimizing only the scene-specific 3D feature volume, while keeping the parameters of the rendering network fixed. Since the feature volumes are independent of the rendering model, we can manipulate and combine scenes by editing their corresponding feature volumes. The edited volume can then be plugged into the rendering model to synthesize high-quality novel views. We demonstrate scene manipulations including: scene mixing; applying rigid and non-rigid transformations; inserting, moving and deleting objects in a scene; while producing photo-realistic novel-view synthesis results.

Most methods for conditional video synthesis use a single modality as the condition. This comes with major limitations. For example, it is problematic for a model conditioned on an image to generate a specific motion trajectory desired by the user since there is no means to provide motion information. Conversely, language information can describe the desired motion, while not precisely defining the content of the video. This work presents a multimodal video generation framework that benefits from text and images provided jointly or separately. We leverage the recent progress in quantized representations for videos and apply a bidirectional transformer with multiple modalities as inputs to predict a discrete video representation. To improve video quality and consistency, we propose a new video token trained with self-learning and an improved mask-prediction algorithm for sampling video tokens. We introduce text augmentation to improve the robustness of the textual representation and diversity of generated videos. Our framework can incorporate various visual modalities, such as segmentation masks, drawings, and partially occluded images. It can generate much longer sequences than the one used for training. In addition, our model can extract visual information as suggested by the text prompt, e.g., “an object in image one is moving northeast”, and generate corresponding videos. We run evaluations on three public datasets and a newly collected dataset labeled with facial attributes, achieving state-of-the-art generation results on all four11Code: https://github.com/snap-research/MMVID and Webpage..

We present a novel method to acquire object representations from online image collections, capturing high-quality geometry and material properties of arbitrary objects from photographs with varying cameras, illumination, and backgrounds. This enables various object-centric rendering applications such as novel-view synthesis, relighting, and harmonized background composition from challenging in-the-wild input. Using a multi-stage approach extending neural radiance fields, we first infer the surface geometry and refine the coarsely estimated initial camera parameters, while leveraging coarse foreground object masks to improve the training efficiency and geometry quality. We also introduce a robust normal estimation technique which eliminates the effect of geometric noise while retaining crucial details. Lastly, we extract surface material properties and ambient illumination, represented in spherical harmonics with extensions that handle transient elements, e.g. sharp shadows. The union of these components results in a highly modular and efficient object acquisition framework. Extensive evaluations and comparisons demonstrate the advantages of our approach in capturing high-quality geometry and appearance properties useful for rendering applications.

Human motion retargeting aims to transfer the motion of one person in a "driving" video or set of images to another person. Existing efforts leverage a long training video from each target person to train a subject-specific motion transfer model. However, the scalability of such methods is limited, as each model can only generate videos for the given target subject, and such training videos are labor-intensive to acquire and process. Few-shot motion transfer techniques, which only require one or a few images from a target, have recently drawn considerable attention. Methods addressing this task generally use either 2D or explicit 3D representations to transfer motion, and in doing so, sacrifice either accurate geometric modeling or the flexibility of an end-to-end learned representation. Inspired by the Transformable Bottleneck Network, which renders novel views and manipulations of rigid objects, we propose an approach based on an implicit volumetric representation of the image content, which can then be spatially manipulated using volumetric flow fields. We address the challenging question of how to aggregate information across different body poses, learning flow fields that allow for combining content from the appropriate regions of input images of highly non-rigid human subjects performing complex motions into a single implicit volumetric representation. This allows us to learn our 3D representation solely from videos of moving people. Armed with both 3D object understanding and end-to-end learned rendering, this categorically novel representation delivers state-of-the-art image generation quality, as shown by our quantitative and qualitative evaluations.

We present an interactive approach to synthesizing realistic variations in facial hair in images, ranging from subtle edits to existing hair to the addition of complex and challenging hair in images of clean-shaven subjects. To circumvent the tedious and computationally expensive tasks of modeling, rendering and compositing the 3D geometry of the target hairstyle using the traditional graphics pipeline, we employ a neural network pipeline that synthesizes realistic and detailed images of facial hair directly in the target image in under one second. The synthesis is controlled by simple and sparse guide strokes from the user defining the general structural and color properties of the target hairstyle. We qualitatively and quantitatively evaluate our chosen method compared to several alternative approaches. We show compelling interactive editing results with a prototype user interface that allows novice users to progressively refine the generated image to match their desired hairstyle, and demonstrate that our approach also allows for flexible and high-fidelity scalp hair synthesis.

We propose a novel approach to performing fine-grained 3D manipulation of image content via a convolutional neural network, which we call the Transformable Bottleneck Network (TBN). It applies given spatial transformations directly to a volumetric bottleneck within our encoder-bottleneck-decoder architecture. Multi-view supervision encourages the network to learn to spatially disentangle the feature space within the bottleneck. The resulting spatial structure can be manipulated with arbitrary spatial transformations. We demonstrate the efficacy of TBNs for novel view synthesis, achieving state-of-the-art results on a challenging benchmark. We demonstrate that the bottlenecks produced by networks trained for this task contain meaningful spatial structure that allows us to intuitively perform a variety of image manipulations in 3D, well beyond the rigid transformations seen during training. These manipulations include non-uniform scaling, non-rigid warping, and combining content from different images. Finally, we extract explicit 3D structure from the bottleneck, performing impressive 3D reconstruction from a single input image.

We present a deep learning-based technique to infer high-quality facial reflectance and geometry given a single unconstrained image of the subject, which may contain partial occlusions and arbitrary illumination conditions. The reconstructed high-resolution textures, which are generated in only a few seconds, include high-resolution skin surface reflectance maps, representing both the diffuse and specular albedo, and medium- and high-frequency displacement maps, thereby allowing us to render compelling digital avatars under novel lighting conditions. To extract this data, we train our deep neural networks with a high-quality skin reflectance and geometry database created with a state-of-the-art multi-view photometric stereo system using polarized gradient illumination. Given the raw facial texture map extracted from the input image, our neural networks synthesize complete reflectance and displacement maps, as well as complete missing regions caused by occlusions. The completed textures exhibit consistent quality throughout the face due to our network architecture, which propagates texture features from the visible region, resulting in high-fidelity details that are consistent with those seen in visible regions. We describe how this highly underconstrained problem is made tractable by dividing the full inference into smaller tasks, which are addressed by dedicated neural networks. We demonstrate the effectiveness of our network design with robust texture completion from images of faces that are largely occluded. With the inferred reflectance and geometry data, we demonstrate the rendering of high-fidelity 3D avatars from a variety of subjects captured under different lighting conditions. In addition, we perform evaluations demonstrating that our method can infer plausible facial reflectance and geometric details comparable to those obtained from high-end capture devices, and outperform alternative approaches that require only a single unconstrained input image.

We present a novel method to realistically puppeteer and animate a face from a single RGB image using a source video sequence. We begin by fitting a multilinear PCA model to obtain the 3D geometry and a single texture of the target face. In order for the animation to be realistic, however, we need dynamic per-frame textures that capture subtle wrinkles and deformations corresponding to the animated facial expressions. This problem is highly underconstrained, as dynamic textures cannot be obtained directly from a single image. Furthermore, if the target face has a closed mouth, it is not possible to obtain actual images of the mouth interior. To address this issue, we train a Deep Generative Network that can infer realistic per-frame texture deformations, including the mouth interior, of the target identity using the per-frame source textures and the single target texture. By retargeting the PCA expression geometry from the source, as well as using the newly inferred texture, we can both animate the face and perform video face replacement on the source video using the target appearance.

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.

Significant challenges currently prohibit expressive interaction in virtual reality (VR). Occlusions introduced by head-mounted displays (HMDs) make existing facial tracking techniques intractable, and even state-of-the-art techniques used for real-time facial tracking in unconstrained environments fail to capture subtle details of the user's facial expressions that are essential for compelling speech animation. We introduce a novel system for HMD users to control a digital avatar in real-time while producing plausible speech animation and emotional expressions. Using a monocular camera attached to an HMD, we record multiple subjects performing various facial expressions and speaking several phonetically-balanced sentences. These images are used with artist-generated animation data corresponding to these sequences to train a convolutional neural network (CNN) to regress images of a user's mouth region to the parameters that control a digital avatar. To make training this system more tractable, we use audio-based alignment techniques to map images of multiple users making the same utterance to the corresponding animation parameters. We demonstrate that this approach is also feasible for tracking the expressions around the user's eye region with an internal infrared (IR) camera, thereby enabling full facial tracking. This system requires no user-specific calibration, uses easily obtainable consumer hardware, and produces high-quality animations of speech and emotional expressions. Finally, we demonstrate the quality of our system on a variety of subjects and evaluate its performance against state-of-the-art real-time facial tracking techniques.

There are currently no solutions for enabling direct face-to-face interaction between virtual reality (VR) users wearing head-mounted displays (HMDs). The main challenge is that the headset obstructs a significant portion of a user's face, preventing effective facial capture with traditional techniques. To advance virtual reality as a next-generation communication platform, we develop a novel HMD that enables 3D facial performance-driven animation in real-time. Our wearable system uses ultra-thin flexible electronic materials that are mounted on the foam liner of the headset to measure surface strain signals corresponding to upper face expressions. These strain signals are combined with a head-mounted RGB-D camera to enhance the tracking in the mouth region and to account for inaccurate HMD placement. To map the input signals to a 3D face model, we perform a single-instance offline training session for each person. For reusable and accurate online operation, we propose a short calibration step to readjust the Gaussian mixture distribution of the mapping before each use. The resulting animations are visually on par with cutting-edge depth sensor-driven facial performance capture systems and hence, are suitable for social interactions in virtual worlds.