Gaussian Splatting has changed the game for real-time photo-realistic rendering. One of the most popular applications of Gaussian Splatting is to create animatable avatars, known as Gaussian Avatars. Recent works have pushed the boundaries of quality and rendering efficiency but suffer from two main limitations. Either they require expensive multi-camera rigs to produce avatars with free-viewpoint rendering, or they can be trained with a single camera but only rendered at high quality from this fixed viewpoint. An ideal model would be trained using a short monocular video or image from available hardware, such as a webcam, and rendered from any view. To this end, we propose GASP: Gaussian Avatars with Synthetic Priors. To overcome the limitations of existing datasets, we exploit the pixel-perfect nature of synthetic data to train a Gaussian Avatar prior. By fitting this prior model to a single photo or video and fine-tuning it, we get a high-quality Gaussian Avatar, which supports 360° rendering. Our prior is only required for fitting, not inference, enabling real-time applications. Through our method, we obtain high-quality, animatable Avatars from limited data which can be animated and rendered at 70fps on commercial hardware.

Avatars are edging into mainstream videoconferencing, but evaluation of how avatar animation modalities contribute to work meeting outcomes has been limited. We report a within-group videoconferencing experiment in which 68 employees of a global technology company, in 16 groups, used the same stylized avatars in three modalities (static picture, audio-animation, and webcam-animation) to complete collaborative decision-making tasks. Quantitatively, for meeting outcomes, webcam-animated avatars improved meeting effectiveness over the picture modality and were also reported to be more comfortable and inclusive than both other modalities. In terms of avatar satisfaction, there was a similar preference for webcam animation as compared to both other modalities. Our qualitative analysis shows participants expressing a preference for the holistic motion of webcam animation, and that meaningful movement outweighs realism for meeting outcomes, as evidenced through a systematic overview of ten thematic factors. We discuss implications for research and commercial deployment and conclude that webcam-animated avatars are a plausible alternative to video in work meetings.

Our work addresses the problem of egocentric human pose estimation from downwards-facing cameras on head-mounted devices (HMD). This presents a challenging scenario, as parts of the body often fall outside of the image or are occluded. Previous solutions minimize this problem by using fish-eye camera lenses to capture a wider view, but these can present hardware design issues. They also predict 2D heat-maps per joint and lift them to 3D space to deal with self-occlusions, but this requires large network architectures which are impractical to deploy on resource-constrained HMDs. We predict pose from images captured with conventional rectilinear camera lenses. This resolves hardware design issues, but means body parts are often out of frame. As such, we directly regress probabilistic joint rotations represented as matrix Fisher distributions for a parameterized body model. This allows us to quantify pose uncertainties and explain out-of-frame or occluded joints. This also removes the need to compute 2D heat-maps and allows for simplified DNN architectures which require less compute. Given the lack of egocentric datasets using rectilinear camera lenses, we introduce the SynthEgo dataset1, a synthetic dataset with 60K stereo images containing high diversity of pose, shape, clothing and skin tone. Our approach achieves state-of-the-art results for this challenging configuration, reducing mean per-joint position error by 23% overall and 58% for the lower body. Our architecture also has eight times fewer parameters and runs twice as fast as the current state-of-the-art. Experiments show that training on our synthetic dataset leads to good generalization to real world images without fine-tuning.1Available at https://microsoft.github.io/SimpleEgo.

We present a method for prediction of a person's hairstyle from a single image. Despite growing use cases in user digitization and enrollment for virtual experiences, available methods are limited, particularly in the range of hairstyles they can capture. Human hair is extremely diverse and lacks any universally accepted description or categorization, making this a challenging task. Most current methods rely on parametric models of hair at a strand level. These approaches, while very promising, are not yet able to represent short, frizzy, coily hair and gathered hairstyles. We instead choose a classification approach which can represent the diversity of hairstyles required for a truly robust and inclusive system. Previous classification approaches have been restricted by poorly labeled data that lacks diversity, imposing constraints on the usefulness of any resulting enrollment system. We use only synthetic data to train our models. This allows for explicit control of diversity of hairstyle attributes, hair colors, facial appearance, poses, environments and other parameters. It also produces noise-free ground-truth labels. We introduce a novel hairstyle taxonomy developed in collaboration with a diverse group of domain experts which we use to balance our training data, supervise our model, and directly measure fairness. We annotate our synthetic training data and a real evaluation dataset using this taxonomy and release both to enable comparison of future hairstyle prediction approaches. We employ an architecture based on a pre-trained feature extraction network in order to improve generalization of our method to real data and predict taxonomy attributes as an auxiliary task to improve accuracy. Results show our method to be significantly more robust for challenging hairstyles than recent parametric approaches.

We tackle the problem of highly-accurate, holistic performance capture for the face, body and hands simultaneously. Motion-capture technologies used in film and game production typically focus only on face, body or hand capture independently, involve complex and expensive hardware and a high degree of manual intervention from skilled operators. While machine-learning-based approaches exist to overcome these problems, they usually only support a single camera, often operate on a single part of the body, do not produce precise world-space results, and rarely generalize outside specific contexts. In this work, we introduce the first technique for markerfree, high-quality reconstruction of the complete human body, including eyes and tongue, without requiring any calibration, manual intervention or custom hardware. Our approach produces stable world-space results from arbitrary camera rigs as well as supporting varied capture environments and clothing. We achieve this through a hybrid approach that leverages machine learning models trained exclusively on synthetic data and powerful parametric models of human shape and motion. We evaluate our method on a number of body, face and hand reconstruction benchmarks and demonstrate state-of-the-art results that generalize on diverse datasets.

State-of-the-art face recognition models show impressive accuracy, achieving over 99.8% on Labeled Faces in the Wild (LFW) dataset. Such models are trained on large-scale datasets that contain millions of real human face images collected from the internet. Web-crawled face images are severely biased (in terms of race, lighting, makeup, etc) and often contain label noise. More importantly, the face images are collected without explicit consent, raising ethical concerns. To avoid such problems, we introduce a large-scale synthetic dataset for face recognition, obtained by rendering digital faces using a computer graphics pipeline1. We first demonstrate that aggressive data augmentation can significantly reduce the synthetic-to-real domain gap. Having full control over the rendering pipeline, we also study how each attribute (e.g., variation in facial pose, accessories and textures) affects the accuracy. Compared to Syn-Face, a recent method trained on GAN-generated synthetic faces, we reduce the error rate on LFW by 52.5% (accuracy from 91.93% to 96.17%). By fine-tuning the network on a smaller number of real face images that could reason-ably be obtained with consent, we achieve accuracy that is comparable to the methods trained on millions of real face images.

Recent advances in synthesizing realistic faces have shown that synthetic training data can replace real data for various face-related computer vision tasks. A question arises: how important is realism? Is the pursuit of photorealism excessive? In this work, we show otherwise. We boost the realism of our synthetic faces by introducing dynamic skin wrinkles in response to facial expressions, and observe significant performance improvements in downstream computer vision tasks. Previous approaches for producing such wrinkles either required prohibitive artist effort to scale across identities and expressions, or were not capable of reconstructing high-frequency skin details with sufficient fidelity. Our key contribution is an approach that produces realistic wrinkles across a large and diverse population of digital humans. Concretely, we formalize the concept of mesh-tension and use it to aggregate possible wrinkles from high-quality expression scans into albedo and displacement texture maps. At synthesis, we use these maps to produce wrinkles even for expressions not represented in the source scans. Additionally, to provide a more nuanced indicator of model performance under deformations resulting from com-pressed expressions, we introduce the 300W-winks evaluation subset and the Pexels dataset of closed eyes and winks.

We demonstrate that it is possible to perform face-related computer vision in the wild using synthetic data alone. The community has long enjoyed the benefits of synthesizing training data with graphics, but the domain gap between real and synthetic data has remained a problem, especially for human faces. Researchers have tried to bridge this gap with data mixing, domain adaptation, and domain-adversarial training, but we show that it is possible to synthesize data with minimal domain gap, so that models trained on synthetic data generalize to real in-the-wild datasets. We describe how to combine a procedurally-generated parametric 3D face model with a comprehensive library of hand-crafted assets to render training images with unprecedented realism and diversity. We train machine learning systems for face-related tasks such as landmark localization and face parsing, showing that synthetic data can both match real data in accuracy as well as open up new approaches where manual labeling would be impossible.

Landmark localisation is a vital step in automatic analysis of facial expressions of animals. Head motion is one of the most challenging problems for face alignment for humans and animals. For animals this is exacerbated by the increased amounts of self-occlusion resulting from variations in head pose. In this paper, we present a novel model for detection of an extensive set of facial landmarks for sheep. A dataset of 850 sheep facial images, annotated with a 25 facial landmark scheme and occlusion information, is introduced: the Sheep Facial Landmarks in the Wild (SFLW) dataset, including a wide range of variations in head-pose and occlusion. Data augmentation techniques are introduced using thin-plate-spline warping and negatively correlated augmentation to boost representation of extreme head poses. We then present a novel pose-informed landmark localisation method based on a fine-tuned CNN model for human head pose estimation. This method is shown to significantly outperform the existing state-of-the-art approach on the introduced SFLW dataset and the viability of the technique for real-world use is demonstrated through the implementation of a near real-time video pipeline.

We introduce the Autonomous Vehicle Acceptance Model (AVAM), a model of user acceptance for autonomous vehicles, adapted from existing models of user acceptance for generic technologies. A 26-item questionnaire is developed in accordance with the model and a survey conducted to evaluate 6 autonomy scenarios. In a pilot survey (n = 54) and follow-up survey (n = 187), the AVAM presented good internal consistency and replicated patterns from previous surveys. Results showed that users were less accepting of high autonomy levels and displayed significantly lower intention to use highly autonomous vehicles. We also assess expected driving engagement of hands, feet and eyes which are shown to be lower for full autonomy compared with all other autonomy levels. This highlighted that partial autonomy, regardless of level, is perceived to require uniformly higher driver engagement than full autonomy. These results can inform experts regarding public perception of autonomy across SAE levels. The AVAM and associated questionnaire enable standardised evaluation of AVs across studies, allowing for meaningful assessment of changes in perception over time and between different technologies.