The training of high-quality, robust machine learning models for speech-driven 3D facial animation requires a large, diverse dataset of high-quality audio-animation pairs. To overcome the lack of such a dataset, recent work has introduced large pre-trained speech encoders that are robust to variations in the input audio and, therefore, enable the facial animation model to generalize across speakers, audio quality, and languages. However, the resulting facial animation models are prohibitively large and lend themselves only to offline inference on a dedicated machine. In this work, we explore on-device, real-time facial animation models in the context of game development. We overcome the lack of large datasets by using hybrid knowledge distillation with pseudo-labeling. Given a large audio dataset, we employ a high-performing teacher model to train very small student models. In contrast to the pre-trained speech encoders, our student models only consist of convolutional and fully-connected layers, removing the need for attention context or recurrent updates. In our experiments, we demonstrate that we can reduce the memory footprint to up to 3.4 MB and required future audio context to up to 81 ms while maintaining high-quality animations. This paves the way for on-device inference, an important step towards realistic, model-driven digital characters.

Abstract Local model-agnostic additive explanation techniques decompose the predicted output of a black-box model into additive feature importance scores. Questions have been raised about the accuracy of the produced local additive explanations. We investigate this by studying whether some of the most popular explanation techniques can accurately explain the decisions of linear additive models. We show that even though the explanations generated by these techniques are linear additives, they can fail to provide accurate explanations when explaining linear additive models. In the experiments, we measure the accuracy of additive explanations, as produced by, e.g., LIME and SHAP, along with the non-additive explanations of Local Permutation Importance (LPI) when explaining Linear and Logistic Regression and Gaussian naive Bayes models over 40 tabular datasets. We also investigate the degree to which different factors, such as the number of numerical or categorical or correlated features, the predictive performance of the black-box model, explanation sample size, similarity metric, and the pre-processing technique used on the dataset can directly affect the accuracy of local explanations.

We consider a setting where users interact with a collection of N items on an online platform. We are given class labels possibly corrupted by noise, and we seek to recover the true class of each item. We postulate a simple probabilistic model of the interactions between users and items, based on the assumption that users interact with classes in different proportions. We then develop a message-passing algorithm that decodes the noisy class labels efficiently. Under suitable assumptions, our method provably recovers all items’ true classes in the large N limit, even when the interaction graph remains sparse. Empirically, we show that our approach is effective on several practical applications, including predicting the location of businesses, the category of consumer goods, and the language of audio content.

Platform ecosystems have witnessed an explosive growth by facilitating interactions between consumers and suppliers. Search systems powering such platforms play an important role in surfacing content in front of users. To maintain a healthy, sustainable platform, systems designers often need to explicitly consider exposing under-served content to users, content which might otherwise remain undiscovered. In this work, we consider the question when we might surface under-served content in search results, and investigate ways to provide exposure to certain content groups. We propose a framework to develop query understanding techniques to identify potential non-focused search queries on a music streaming platform, where users' information needs are non-specific enough to expose under-served content without severely impacting user satisfaction. We present insights from a search ranker deployed at scale and present results from live A/B test targeting a random sample of 72 million users and 593 million sessions, to compare performance of different methods considered to identify non-focused queries for surfacing under-served content.

Auto encoding models have been extensively studied in recent years. They provide an efficient framework for sample generation, as well as for analysing feature learning. Furthermore, they are efficient in performing interpolations between data-points in semantically meaningful ways. In this paper, we introduce a method for generating sequence samples from auto encoders trained on flattened sequences (e.g video sample from auto encoders trained on single frames); as well as a canonical, dimension independent method for generating stochastic interpolations. The distribution of interpolation paths is represented as the distribution of a bridge process constructed from an artificial random data generating process in the latent space, having the prior distribution as its invariant distribution.

Many modern unsupervised or semi-supervised machine learning algorithms rely on Bayesian probabilistic models. These models are usually intractable and thus require approximate inference. Variational inference (VI) lets us approximate a high-dimensional Bayesian posterior with a simpler variational distribution by solving an optimization problem. This approach has been successfully applied to various models and large-scale applications. In this review, we give an overview of recent trends in variational inference. We first introduce standard mean field variational inference, then review recent advances focusing on the following aspects: (a) scalable VI, which includes stochastic approximations, (b) generic VI, which extends the applicability of VI to a large class of otherwise intractable models, such as non-conjugate models, (c) accurate VI, which includes variational models beyond the mean field approximation or with atypical divergences, and (d) amortized VI, which implements the inference over local latent variables with inference networks. Finally, we provide a summary of promising future research directions.

Fluent and safe interactions of humans and robots require both partners to anticipate the others' actions. The bottleneck of most methods is the lack of an accurate model of natural human motion. In this work, we present a conditional variational autoencoder that is trained to predict a window of future human motion given a window of past frames. Using skeletal data obtained from RGB depth images, we show how this unsupervised approach can be used for online motion prediction for up to 1660 ms. Additionally, we demonstrate online target prediction within the first 300-500 ms after motion onset without the use of target specific training data. The advantage of our probabilistic approach is the possibility to draw samples of possible future motion patterns. Finally, we investigate how movements and kinematic cues are represented on the learned low dimensional manifold.

Generative models of 3D human motion are often restricted to a small number of activities and can therefore not generalize well to novel movements or applications. In this work we propose a deep learning framework for human motion capture data that learns a generic representation from a large corpus of motion capture data and generalizes well to new, unseen, motions. Using an encoding-decoding network that learns to predict future 3D poses from the most recent past, we extract a feature representation of human motion. Most work on deep learning for sequence prediction focuses on video and speech. Since skeletal data has a different structure, we present and evaluate different network architectures that make different assumptions about time dependencies and limb correlations. To quantify the learned features, we use the output of different layers for action classification and visualize the receptive fields of the network units. Our method outperforms the recent state of the art in skeletal motion prediction even though these use action specific training data. Our results show that deep feedforward networks, trained from a generic mocap database, can successfully be used for feature extraction from human motion data and that this representation can be used as a foundation for classification and prediction.

We present a general framework to autonomously achieve the task of finding a sequence of actions that result in a desired state. Autonomy is acquired by learning sensorimotor patterns of a robot, while it is interacting with its environment.

Modeling of physical human-robot collaborations is generally a challenging problem due to the unpredictive nature of human behavior. To address this issue, we present a data-efficient reinforcement learning framework which enables a robot to learn how to collaborate with a human partner. The robot learns the task from its own sensorimotor experiences in an unsupervised manner. The uncertainty in the interaction is modeled using Gaussian processes (GP) to implement a forward model and an action-value function. Optimal action selection given the uncertain GP model is ensured by Bayesian optimization. We apply the framework to a scenario in which a human and a PR2 robot jointly control the ball position on a plank based on vision and force/torque data. Our experimental results show the suitability of the proposed method in terms of fast and data-efficient model learning, optimal action selection under uncertainty and equal role sharing between the partners.