Capturing and editing full head performances enables the creation of virtual characters with various applications such as extended reality and media production. The past few years witnessed a steep rise in the photorealism of human head avatars. Such avatars can be controlled through different input data modalities, including RGB, audio, depth, IMUs and others. While these data modalities provide effective means of control, they mostly focus on editing the head movements such as the facial expressions, head pose and/or camera viewpoint. In this paper, we propose AvatarStudio, a text-based method for editing the appearance of a dynamic full head avatar. Our approach builds on existing work to capture dynamic performances of human heads using neural radiance field (NeRF) and edits this representation with a text-to-image diffusion model. Specifically, we introduce an optimization strategy for incorporating multiple keyframes representing different camera viewpoints and time stamps of a video performance into a single diffusion model. Using this personalized diffusion model, we edit the dynamic NeRF by introducing view-and-time-aware Score Distillation Sampling (VT-SDS) following a model-based guidance approach. Our method edits the full head in a canonical space, and then propagates these edits to remaining time steps via a pretrained deformation network. We evaluate our method visually and numerically via a user study, and results show that our method outperforms existing approaches. Our experiments validate the design choices of our method and highlight that our edits are genuine, personalized, as well as 3D- and time-consistent.
Advances in 3D-aware generative models have pushed the boundary of image synthesis with explicit camera control. To achieve high-resolution image synthesis, several attempts have been made to design efficient generators, such as hybrid architectures with both 3D and 2D components. However, such a design compromises multiview consistency, and the design of a pure 3D generator with high resolution is still an open problem. In this work, we present Generative Volumetric Primitives (GVP), the first pure 3D generative model that can sample and render 512-resolution images in real-time. GVP jointly models a number of volumetric primitives and their spatial information, both of which can be efficiently generated via a 2D convolutional network. The mixture of these primitives naturally captures the sparsity and correspondence in the 3D volume. The training of such a generator with a high degree of freedom is made possible through a knowledge distillation technique. Experiments on several datasets demonstrate superior efficiency and 3D consistency of GVP over the state-of-the-art.
Multi-view volumetric rendering techniques have recently shown great potential in modeling and synthesizing high-quality head avatars. A common approach to capture full head dynamic performances is to track the underlying geometry using a mesh-based template or 3D cube-based graphics primitives. While these model-based approaches achieve promising results, they often fail to learn complex geometric details such as the mouth interior, hair, and topological changes over time. This paper presents a novel approach to building highly photorealistic digital head avatars. Our method learns a canonical space via an implicit function parameterized by a neural network. It leverages multiresolution hash encoding in the learned feature space, allowing for high-quality, faster training and high-resolution rendering. At test time, our method is driven by a monocular RGB video. Here, an image encoder extracts face-specific features that also condition the learnable canonical space. This encourages deformation-dependent texture variations during training. We also propose a novel optical flow based loss that ensures correspondences in the learned canonical space, thus encouraging artifact-free and temporally consistent renderings. We show results on challenging facial expressions and show free-viewpoint renderings at interactive real-time rates for medium image resolutions. Our method outperforms all existing approaches, both visually and numerically. We will release our multiple-identity dataset to encourage further research. Our Project page is available at: https://vcai.mpi-inf.mpg.de/projects/HQ3DAvatar/
3D Morphable models of the human body capture variations among subjects and are useful in reconstruction and editing applications. Current dental models use an explicit mesh scene representation and model only the teeth, ignoring the gum. In this work, we present the first parametric 3D morphable dental model for both teeth and gum. Our model uses an implicit scene representation and is learned from rigidly aligned scans. It is based on a component-wise representation for each tooth and the gum, together with a learnable latent code for each of such components. It also learns a template shape thus enabling several applications such as segmentation, interpolation, and tooth replacement. Our reconstruction quality is on par with the most advanced global implicit representations while enabling novel applications. Project page: https://vcai.mpi-inf.mpg.de/projects/DMM/
3D generative models of objects enable photorealistic image synthesis with 3D control. Existing methods model the scene as a global scene representation, ignoring the compositional aspect of the scene. Compositional reasoning can enable a wide variety of editing applications, in addition to enabling generalizable 3D reasoning. In this paper, we present a compositional generative model, where each semantic part of the object is represented as an independent 3D representation learned from only in-the-wild 2D data. We start with a global generative model (GAN) and learn to decompose it into different semantic parts using supervision from 2D segmentation masks. We then learn to composite independently sampled parts in order to create coherent global scenes. Different parts can be independently sampled while keeping the rest of the object fixed. We evaluate our method on a wide variety of objects and parts and demonstrate editing applications.
Learning coordinate-based volumetric 3D scene representations such as neural radiance fields (NeRF) has been so far studied assuming RGB or RGB-D images as inputs. At the same time, it is known from the neuroscience literature that human visual system (HVS) is tailored to process asynchronous brightness changes rather than synchronous RGB images, in order to build and continuously update mental 3D representations of the surroundings for navigation and survival. Visual sensors that were inspired by HVS principles are event cameras. Thus, events are sparse and asynchronous per-pixel brightness (or colour channel) change signals. In contrast to existing works on neural 3D scene representation learning, this paper approaches the problem from a new perspective. We demonstrate that it is possible to learn NeRF suitable for novel-view synthesis in the RGB space from asynchronous event streams. Our models achieve high visual accuracy of the rendered novel views of challenging scenes in the RGB space, even though they are trained with substantially fewer data (i.e., event streams from a single event camera moving around the object) and more efficiently (due to the inherent sparsity of event streams) than the existing NeRF models trained with RGB images. We will release our datasets and the source code, see https://4dqv.mpi-inf.mpg.de/EventNeRF/.
Shape-from-Template (SfT) methods estimate 3D surface deformations from a single monocular RGB camera while assuming a 3D state known in advance (a template). This is an important yet challenging problem due to the under-constrained nature of the monocular setting. Existing SfT techniques predominantly use geometric and simplified deformation models, which often limits their reconstruction abilities. In contrast to previous works, this paper proposes a new SfT approach explaining 2D observations through physical simulations accounting for forces and material properties. Our differentiable physics simulator regularises the surface evolution and optimises the material elastic properties such as bending coefficients, stretching stiffness and density. We use a differentiable renderer to minimise the dense reprojection error between the estimated 3D states and the input images and recover the deformation parameters using an adaptive gradient-based optimisation. For the evaluation, we record with an RGB-D camera challenging real surfaces exposed to physical forces with various material properties and textures. Our approach significantly reduces the 3D reconstruction error compared to multiple competing methods. For the source code and data, see https://4dqv.mpi-inf.mpg.de/phi-SfT/.
Photorealistic editing of outdoor scenes from photographs requires a profound understanding of the image formation process and an accurate estimation of the scene geometry, reflectance and illumination. A delicate manipulation of the lighting can then be performed while keeping the scene albedo and geometry unaltered. We present NeRF-OSR, i.e., the first approach for outdoor scene relighting based on neural radiance fields. In contrast to the prior art, our technique allows simultaneous editing of both scene illumination and camera viewpoint using only a collection of outdoor photos shot in uncontrolled settings. Moreover, it enables direct control over the scene illumination, as defined through a spherical harmonics model. It also includes a dedicated network for shadow reproduction, which is crucial for high-quality outdoor scene relighting. To evaluate the proposed method, we collect a new benchmark dataset of several outdoor sites, where each site is photographed from multiple viewpoints and at different timings. For each timing, a 360 degrees environment map is captured together with a colour-calibration chequerboard to allow accurate numerical evaluations on real data against ground truth. Comparisons against state of the art show that NeRF-OSR enables controllable lighting and viewpoint editing at higher quality and with realistic self-shadowing reproduction. Our method and the dataset will be made publicly available at https://4dqv.mpi-inf.mpg.de/NeRF-OSR/.
Generative adversarial models (GANs) continue to produce advances in terms of the visual quality of still images, as well as the learning of temporal correlations. However, few works manage to combine these two interesting capabilities for the synthesis of video content: Most methods require an extensive training dataset in order to learn temporal correlations, while being rather limited in the resolution and visual quality of their output frames. In this paper, we present a novel approach to the video synthesis problem that helps to greatly improve visual quality and drastically reduce the amount of training data and resources necessary for generating video content. Our formulation separates the spatial domain, in which individual frames are synthesized, from the temporal domain, in which motion is generated. For the spatial domain we make use of a pre-trained StyleGAN network, the latent space of which allows control over the appearance of the objects it was trained for. The expressive power of this model allows us to embed our training videos in the StyleGAN latent space. Our temporal architecture is then trained not on sequences of RGB frames, but on sequences of StyleGAN latent codes. The advantageous properties of the StyleGAN space simplify the discovery of temporal correlations. We demonstrate that it suffices to train our temporal architecture on only 10 minutes of footage of 1 subject for about 6 hours. After training, our model can not only generate new portrait videos for the training subject, but also for any random subject which can be embedded in the StyleGAN space.
We introduce a method for egocentric videoconferencing that enables hands-free video calls, for instance by people wearing smart glasses or other mixed-reality devices. Videoconferencing portrays valuable non-verbal communication and face expression cues, but usually requires a front-facing camera. Using a frontal camera in a hands-free setting when a person is on the move is impractical. Even holding a mobile phone camera in the front of the face while sitting for a long duration is not convenient. To overcome these issues, we propose a low-cost wearable egocentric camera setup that can be integrated into smart glasses. Our goal is to mimic a classical video call, and therefore, we transform the egocentric perspective of this camera into a front facing video. To this end, we employ a conditional generative adversarial neural network that learns a transition from the highly distorted egocentric views to frontal views common in videoconferencing. Our approach learns to transfer expression details directly from the egocentric view without using a complex intermediate parametric expressions model, as it is used by related face reenactment methods. We successfully handle subtle expressions, not easily captured by parametric blendshape-based solutions, e.g., tongue movement, eye movements, eye blinking, strong expressions and depth varying movements. To get control over the rigid head movements in the target view, we condition the generator on synthetic renderings of a moving neutral face. This allows us to synthesis results at different head poses. Our technique produces temporally smooth video-realistic renderings in real-time using a video-to-video translation network in conjunction with a temporal discriminator. We demonstrate the improved capabilities of our technique by comparing against related state-of-the art approaches.