Matthew Johnson

Erroll Wood, Tadas Baltrusaitis, Charlie Hewitt et al.

Landmarks often play a key role in face analysis, but many aspects of identity or expression cannot be represented by sparse landmarks alone. Thus, in order to reconstruct faces more accurately, landmarks are often combined with additional signals like depth images or techniques like differentiable rendering. Can we keep things simple by just using more landmarks? In answer, we present the first method that accurately predicts 10x as many landmarks as usual, covering the whole head, including the eyes and teeth. This is accomplished using synthetic training data, which guarantees perfect landmark annotations. By fitting a morphable model to these dense landmarks, we achieve state-of-the-art results for monocular 3D face reconstruction in the wild. We show that dense landmarks are an ideal signal for integrating face shape information across frames by demonstrating accurate and expressive facial performance capture in both monocular and multi-view scenarios. This approach is also highly efficient: we can predict dense landmarks and fit our 3D face model at over 150FPS on a single CPU thread. Please see our website: https://microsoft.github.io/DenseLandmarks/.

Stephan J. Garbin, Marek Kowalski, Virginia Estellers et al.

The recent increase in popularity of volumetric representations for scene reconstruction and novel view synthesis has put renewed focus on animating volumetric content at high visual quality and in real-time. While implicit deformation methods based on learned functions can produce impressive results, they are `black boxes' to artists and content creators, they require large amounts of training data to generalise meaningfully, and they do not produce realistic extrapolations outside the training data. In this work we solve these issues by introducing a volume deformation method which is real-time, easy to edit with off-the-shelf software and can extrapolate convincingly. To demonstrate the versatility of our method, we apply it in two scenarios: physics-based object deformation and telepresence where avatars are controlled using blendshapes. We also perform thorough experiments showing that our method compares favourably to both volumetric approaches combined with implicit deformation and methods based on mesh deformation.

Stanislaw Szymanowicz, Virginia Estellers, Tadas Baltrusaitis et al.

Delivering immersive, 3D experiences for human communication requires a method to obtain 360 degree photo-realistic avatars of humans. To make these experiences accessible to all, only commodity hardware, like mobile phone cameras, should be necessary to capture the data needed for avatar creation. For avatars to be rendered realistically from any viewpoint, we require training images and camera poses from all angles. However, we cannot rely on there being trackable features in the foreground or background of all images for use in estimating poses, especially from the side or back of the head. To overcome this, we propose a novel landmark detector trained on synthetic data to estimate camera poses from 360 degree mobile phone videos of a human head for use in a multi-stage optimization process which creates a photo-realistic avatar. We perform validation experiments with synthetic data and showcase our method on 360 degree avatars trained from mobile phone videos.

Gemini Team, Petko Georgiev, Ving Ian Lei et al. · deepmind, mila

In this report, we introduce the Gemini 1.5 family of models, representing the next generation of highly compute-efficient multimodal models capable of recalling and reasoning over fine-grained information from millions of tokens of context, including multiple long documents and hours of video and audio. The family includes two new models: (1) an updated Gemini 1.5 Pro, which exceeds the February version on the great majority of capabilities and benchmarks; (2) Gemini 1.5 Flash, a more lightweight variant designed for efficiency with minimal regression in quality. Gemini 1.5 models achieve near-perfect recall on long-context retrieval tasks across modalities, improve the state-of-the-art in long-document QA, long-video QA and long-context ASR, and match or surpass Gemini 1.0 Ultra's state-of-the-art performance across a broad set of benchmarks. Studying the limits of Gemini 1.5's long-context ability, we find continued improvement in next-token prediction and near-perfect retrieval (>99%) up to at least 10M tokens, a generational leap over existing models such as Claude 3.0 (200k) and GPT-4 Turbo (128k). Finally, we highlight real-world use cases, such as Gemini 1.5 collaborating with professionals on completing their tasks achieving 26 to 75% time savings across 10 different job categories, as well as surprising new capabilities of large language models at the frontier; when given a grammar manual for Kalamang, a language with fewer than 200 speakers worldwide, the model learns to translate English to Kalamang at a similar level to a person who learned from the same content.

Gheorghe Comanici, Eric Bieber, Mike Schaekermann et al. · amazon-science, baidu

In this report, we introduce the Gemini 2.X model family: Gemini 2.5 Pro and Gemini 2.5 Flash, as well as our earlier Gemini 2.0 Flash and Flash-Lite models. Gemini 2.5 Pro is our most capable model yet, achieving SoTA performance on frontier coding and reasoning benchmarks. In addition to its incredible coding and reasoning skills, Gemini 2.5 Pro is a thinking model that excels at multimodal understanding and it is now able to process up to 3 hours of video content. Its unique combination of long context, multimodal and reasoning capabilities can be combined to unlock new agentic workflows. Gemini 2.5 Flash provides excellent reasoning abilities at a fraction of the compute and latency requirements and Gemini 2.0 Flash and Flash-Lite provide high performance at low latency and cost. Taken together, the Gemini 2.X model generation spans the full Pareto frontier of model capability vs cost, allowing users to explore the boundaries of what is possible with complex agentic problem solving.

Gemini Team, Rohan Anil, Sebastian Borgeaud et al.

This report introduces a new family of multimodal models, Gemini, that exhibit remarkable capabilities across image, audio, video, and text understanding. The Gemini family consists of Ultra, Pro, and Nano sizes, suitable for applications ranging from complex reasoning tasks to on-device memory-constrained use-cases. Evaluation on a broad range of benchmarks shows that our most-capable Gemini Ultra model advances the state of the art in 30 of 32 of these benchmarks - notably being the first model to achieve human-expert performance on the well-studied exam benchmark MMLU, and improving the state of the art in every one of the 20 multimodal benchmarks we examined. We believe that the new capabilities of the Gemini family in cross-modal reasoning and language understanding will enable a wide variety of use cases. We discuss our approach toward post-training and deploying Gemini models responsibly to users through services including Gemini, Gemini Advanced, Google AI Studio, and Cloud Vertex AI.

Ian Perera, Matthew Johnson, Carson Wilber

Natural language analysis of human collaborative chat dialogues is an understudied domain with many unique challenges: a large number of dialogue act labels, underspecified and dynamic tasks, interleaved topics, and long-range contextual dependence. While prior work has studied broad metrics of team dialogue and associated performance using methods such as LSA, there has been little effort in generating fine-grained descriptions of team dynamics and individual performance from dialogue. We describe initial work towards developing an explainable analytics tool in the software development domain using Slack chats mined from our organization, including generation of a novel, hierarchical labeling scheme; design of descriptive metrics based on the frequency of occurrence of dialogue acts; and initial results using a transformer + CRF architecture to incorporate long-range context.

Aidan Clark, Diego de las Casas, Aurelia Guy et al.

The performance of a language model has been shown to be effectively modeled as a power-law in its parameter count. Here we study the scaling behaviors of Routing Networks: architectures that conditionally use only a subset of their parameters while processing an input. For these models, parameter count and computational requirement form two independent axes along which an increase leads to better performance. In this work we derive and justify scaling laws defined on these two variables which generalize those known for standard language models and describe the performance of a wide range of routing architectures trained via three different techniques. Afterwards we provide two applications of these laws: first deriving an Effective Parameter Count along which all models scale at the same rate, and then using the scaling coefficients to give a quantitative comparison of the three routing techniques considered. Our analysis derives from an extensive evaluation of Routing Networks across five orders of magnitude of size, including models with hundreds of experts and hundreds of billions of parameters.

Jack W. Rae, Sebastian Borgeaud, Trevor Cai et al.

Language modelling provides a step towards intelligent communication systems by harnessing large repositories of written human knowledge to better predict and understand the world. In this paper, we present an analysis of Transformer-based language model performance across a wide range of model scales -- from models with tens of millions of parameters up to a 280 billion parameter model called Gopher. These models are evaluated on 152 diverse tasks, achieving state-of-the-art performance across the majority. Gains from scale are largest in areas such as reading comprehension, fact-checking, and the identification of toxic language, but logical and mathematical reasoning see less benefit. We provide a holistic analysis of the training dataset and model's behaviour, covering the intersection of model scale with bias and toxicity. Finally we discuss the application of language models to AI safety and the mitigation of downstream harms.

Erroll Wood, Tadas Baltrušaitis, Charlie Hewitt et al.

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 labelling would be impossible.

Stephan J. Garbin, Marek Kowalski, Matthew Johnson et al.

Recent work on Neural Radiance Fields (NeRF) showed how neural networks can be used to encode complex 3D environments that can be rendered photorealistically from novel viewpoints. Rendering these images is very computationally demanding and recent improvements are still a long way from enabling interactive rates, even on high-end hardware. Motivated by scenarios on mobile and mixed reality devices, we propose FastNeRF, the first NeRF-based system capable of rendering high fidelity photorealistic images at 200Hz on a high-end consumer GPU. The core of our method is a graphics-inspired factorization that allows for (i) compactly caching a deep radiance map at each position in space, (ii) efficiently querying that map using ray directions to estimate the pixel values in the rendered image. Extensive experiments show that the proposed method is 3000 times faster than the original NeRF algorithm and at least an order of magnitude faster than existing work on accelerating NeRF, while maintaining visual quality and extensibility.

Emil Annevelink, Rachel Kurchin, Eric Muckley et al.

Large-scale electrification is vital to addressing the climate crisis, but several scientific and technological challenges remain to fully electrify both the chemical industry and transportation. In both of these areas, new electrochemical materials will be critical, but their development currently relies heavily on human-time-intensive experimental trial and error and computationally expensive first-principles, meso-scale and continuum simulations. We present an automated workflow, AutoMat, that accelerates these computational steps by introducing both automated input generation and management of simulations across scales from first principles to continuum device modeling. Furthermore, we show how to seamlessly integrate multi-fidelity predictions such as machine learning surrogates or automated robotic experiments "in-the-loop". The automated framework is implemented with design space search techniques to dramatically accelerate the overall materials discovery pipeline by implicitly learning design features that optimize device performance across several metrics. We discuss the benefits of AutoMat using examples in electrocatalysis and energy storage and highlight lessons learned.

Tadas Baltrusaitis, Erroll Wood, Virginia Estellers et al.

Analysis of faces is one of the core applications of computer vision, with tasks ranging from landmark alignment, head pose estimation, expression recognition, and face recognition among others. However, building reliable methods requires time-consuming data collection and often even more time-consuming manual annotation, which can be unreliable. In our work we propose synthesizing such facial data, including ground truth annotations that would be almost impossible to acquire through manual annotation at the consistency and scale possible through use of synthetic data. We use a parametric face model together with hand crafted assets which enable us to generate training data with unprecedented quality and diversity (varying shape, texture, expression, pose, lighting, and hair).

Stephan J. Garbin, Marek Kowalski, Matthew Johnson et al.

Generating photorealistic images of human faces at scale remains a prohibitively difficult task using computer graphics approaches. This is because these require the simulation of light to be photorealistic, which in turn requires physically accurate modelling of geometry, materials, and light sources, for both the head and the surrounding scene. Non-photorealistic renders however are increasingly easy to produce. In contrast to computer graphics approaches, generative models learned from more readily available 2D image data have been shown to produce samples of human faces that are hard to distinguish from real data. The process of learning usually corresponds to a loss of control over the shape and appearance of the generated images. For instance, even simple disentangling tasks such as modifying the hair independently of the face, which is trivial to accomplish in a computer graphics approach, remains an open research question. In this work, we propose an algorithm that matches a non-photorealistic, synthetically generated image to a latent vector of a pretrained StyleGAN2 model which, in turn, maps the vector to a photorealistic image of a person of the same pose, expression, hair, and lighting. In contrast to most previous work, we require no synthetic training data. To the best of our knowledge, this is the first algorithm of its kind to work at a resolution of 1K and represents a significant leap forward in visual realism.

Marek Kowalski, Stephan J. Garbin, Virginia Estellers et al.

Our ability to sample realistic natural images, particularly faces, has advanced by leaps and bounds in recent years, yet our ability to exert fine-tuned control over the generative process has lagged behind. If this new technology is to find practical uses, we need to achieve a level of control over generative networks which, without sacrificing realism, is on par with that seen in computer graphics and character animation. To this end we propose ConfigNet, a neural face model that allows for controlling individual aspects of output images in semantically meaningful ways and that is a significant step on the path towards finely-controllable neural rendering. ConfigNet is trained on real face images as well as synthetic face renders. Our novel method uses synthetic data to factorize the latent space into elements that correspond to the inputs of a traditional rendering pipeline, separating aspects such as head pose, facial expression, hair style, illumination, and many others which are very hard to annotate in real data. The real images, which are presented to the network without labels, extend the variety of the generated images and encourage realism. Finally, we propose an evaluation criterion using an attribute detection network combined with a user study and demonstrate state-of-the-art individual control over attributes in the output images.

Andrew Cook, Bappaditya Mandal, Donna Berry et al.

This paper has been withdrawn by the authors due to insufficient or definition error(s) in the ethics approval protocol. Autism spectrum disorders (ASD) impact the cognitive, social, communicative and behavioral abilities of an individual. The development of new clinical decision support systems is of importance in reducing the delay between presentation of symptoms and an accurate diagnosis. In this work, we contribute a new database consisting of video clips of typical (normal) and atypical (such as hand flapping, spinning or rocking) behaviors, displayed in natural settings, which have been collected from the YouTube video website. We propose a preliminary non-intrusive approach based on skeleton keypoint identification using pretrained deep neural networks on human body video clips to extract features and perform body movement analysis that differentiates typical and atypical behaviors of children. Experimental results on the newly contributed database show that our platform performs best with decision tree as the classifier when compared to other popular methodologies and offers a baseline against which alternate approaches may developed and tested.

Dustin Tran, Matthew Hoffman, Dave Moore et al.

We describe a simple, low-level approach for embedding probabilistic programming in a deep learning ecosystem. In particular, we distill probabilistic programming down to a single abstraction---the random variable. Our lightweight implementation in TensorFlow enables numerous applications: a model-parallel variational auto-encoder (VAE) with 2nd-generation tensor processing units (TPUv2s); a data-parallel autoregressive model (Image Transformer) with TPUv2s; and multi-GPU No-U-Turn Sampler (NUTS). For both a state-of-the-art VAE on 64x64 ImageNet and Image Transformer on 256x256 CelebA-HQ, our approach achieves an optimal linear speedup from 1 to 256 TPUv2 chips. With NUTS, we see a 100x speedup on GPUs over Stan and 37x over PyMC3.

Daniel Emaasit, Matthew Johnson

Scientific fields such as insider-threat detection and highway-safety planning often lack sufficient amounts of time-series data to estimate statistical models for the purpose of scientific discovery. Moreover, the available limited data are quite noisy. This presents a major challenge when estimating time-series models that are robust to overfitting and have well-calibrated uncertainty estimates. Most of the current literature in these fields involve visualizing the time-series for noticeable structure and hard coding them into pre-specified parametric functions. This approach is associated with two limitations. First, given that such trends may not be easily noticeable in small data, it is difficult to explicitly incorporate expressive structure into the models during formulation. Second, it is difficult to know $\textit{a priori}$ the most appropriate functional form to use. To address these limitations, a nonparametric Bayesian approach was proposed to implicitly capture hidden structure from time series having limited data. The proposed model, a Gaussian process with a spectral mixture kernel, precludes the need to pre-specify a functional form and hard code trends, is robust to overfitting and has well-calibrated uncertainty estimates.

Ardavan Saeedi, Matthew Hoffman, Matthew Johnson et al.

We propose the segmented iHMM (siHMM), a hierarchical infinite hidden Markov model (iHMM) that supports a simple, efficient inference scheme. The siHMM is well suited to segmentation problems, where the goal is to identify points at which a time series transitions from one relatively stable regime to a new regime. Conventional iHMMs often struggle with such problems, since they have no mechanism for distinguishing between high- and low-level dynamics. Hierarchical HMMs (HHMMs) can do better, but they require much more complex and expensive inference algorithms. The siHMM retains the simplicity and efficiency of the iHMM, but outperforms it on a variety of segmentation problems, achieving performance that matches or exceeds that of a more complicated HHMM.

Mohammad Norouzi, Maxwell D. Collins, Matthew Johnson et al.

Decision trees and randomized forests are widely used in computer vision and machine learning. Standard algorithms for decision tree induction optimize the split functions one node at a time according to some splitting criteria. This greedy procedure often leads to suboptimal trees. In this paper, we present an algorithm for optimizing the split functions at all levels of the tree jointly with the leaf parameters, based on a global objective. We show that the problem of finding optimal linear-combination (oblique) splits for decision trees is related to structured prediction with latent variables, and we formulate a convex-concave upper bound on the tree's empirical loss. The run-time of computing the gradient of the proposed surrogate objective with respect to each training exemplar is quadratic in the the tree depth, and thus training deep trees is feasible. The use of stochastic gradient descent for optimization enables effective training with large datasets. Experiments on several classification benchmarks demonstrate that the resulting non-greedy decision trees outperform greedy decision tree baselines.