Ognjen Rudovic

Vineet Garg, Ognjen Rudovic, Pranay Dighe et al.

We address the problem of detecting speech directed to a device that does not contain a specific wake-word. Specifically, we focus on audio coming from a touch-based invocation. Mitigating virtual assistants (VAs) activation due to accidental button presses is critical for user experience. While the majority of approaches to false trigger mitigation (FTM) are designed to detect the presence of a target keyword, inferring user intent in absence of keyword is difficult. This also poses a challenge when creating the training/evaluation data for such systems due to inherent ambiguity in the user's data. To this end, we propose a novel FTM approach that uses weakly-labeled training data obtained with a newly introduced data sampling strategy. While this sampling strategy reduces data annotation efforts, the data labels are noisy as the data are not annotated manually. We use these data to train an acoustics-only model for the FTM task by regularizing its loss function via knowledge distillation from an ASR-based (LatticeRNN) model. This improves the model decisions, resulting in 66% gain in accuracy, as measured by equal-error-rate (EER), over the base acoustics-only model. We also show that the ensemble of the LatticeRNN and acoustic-distilled models brings further accuracy improvement of 20%.

Ognjen Rudovic, Akanksha Bindal, Vineet Garg et al.

When interacting with smart devices such as mobile phones or wearables, the user typically invokes a virtual assistant (VA) by saying a keyword or by pressing a button on the device. However, in many cases, the VA can accidentally be invoked by the keyword-like speech or accidental button press, which may have implications on user experience and privacy. To this end, we propose an acoustic false-trigger-mitigation (FTM) approach for on-device device-directed speech detection that simultaneously handles the voice-trigger and touch-based invocation. To facilitate the model deployment on-device, we introduce a new streaming decision layer, derived using the notion of temporal convolutional networks (TCN) [1], known for their computational efficiency. To the best of our knowledge, this is the first approach that can detect device-directed speech from more than one invocation type in a streaming fashion. We compare this approach with streaming alternatives based on vanilla Average layer, and canonical LSTMs, and show: (i) that all the models show only a small degradation in accuracy compared with the invocation-specific models, and (ii) that the newly introduced streaming TCN consistently performs better or comparable with the alternatives, while mitigating device undirected speech faster in time, and with (relative) reduction in runtime peak-memory over the LSTM-based approach of 33% vs. 7%, when compared to a non-streaming counterpart.

Zhonghao Shi, Thomas R Groechel, Shomik Jain et al.

Affect-aware socially assistive robotics (SAR) has shown great potential for augmenting interventions for children with autism spectrum disorders (ASD). However, current SAR cannot yet perceive the unique and diverse set of atypical cognitive-affective behaviors from children with ASD in an automatic and personalized fashion in long-term (multi-session) real-world interactions. To bridge this gap, this work designed and validated personalized models of arousal and valence for children with ASD using a multi-session in-home dataset of SAR interventions. By training machine learning (ML) algorithms with supervised domain adaptation (s-DA), the personalized models were able to trade off between the limited individual data and the more abundant less personal data pooled from other study participants. We evaluated the effects of personalization on a long-term multimodal dataset consisting of 4 children with ASD with a total of 19 sessions, and derived inter-rater reliability (IR) scores for binary arousal (IR = 83%) and valence (IR = 81%) labels between human annotators. Our results show that personalized Gradient Boosted Decision Trees (XGBoost) models with s-DA outperformed two non-personalized individualized and generic model baselines not only on the weighted average of all sessions, but also statistically (p < .05) across individual sessions. This work paves the way for the development of personalized autonomous SAR systems tailored toward individuals with atypical cognitive-affective and socio-emotional needs.

Ognjen Rudovic, Nicolas Tobis, Sebastian Kaltwang et al.

Standard machine learning approaches require centralizing the users' data in one computer or a shared database, which raises data privacy and confidentiality concerns. Therefore, limiting central access is important, especially in healthcare settings, where data regulations are strict. A potential approach to tackling this is Federated Learning (FL), which enables multiple parties to collaboratively learn a shared prediction model by using parameters of locally trained models while keeping raw training data locally. In the context of AI-assisted pain-monitoring, we wish to enable confidentiality-preserving and unobtrusive pain estimation for long-term pain-monitoring and reduce the burden on the nursing staff who perform frequent routine check-ups. To this end, we propose a novel Personalized Federated Deep Learning (PFDL) approach for pain estimation from face images. PFDL performs collaborative training of a deep model, implemented using a lightweight CNN architecture, across different clients (i.e., subjects) without sharing their face images. Instead of sharing all parameters of the model, as in standard FL, PFDL retains the last layer locally (used to personalize the pain estimates). This (i) adds another layer of data confidentiality, making it difficult for an adversary to infer pain levels of the target subject, while (ii) personalizing the pain estimation to each subject through local parameter tuning. We show using a publicly available dataset of face videos of pain (UNBC-McMaster Shoulder Pain Database), that PFDL performs comparably or better than the standard centralized and FL algorithms, while further enhancing data privacy. This, has the potential to improve traditional pain monitoring by making it more secure, computationally efficient, and scalable to a large number of individuals (e.g., for in-home pain monitoring), providing timely and unobtrusive pain measurement.

Mihee Lee, Ognjen Rudovic, Vladimir Pavlovic et al.

Detecting facial action units (AU) is one of the fundamental steps in automatic recognition of facial expression of emotions and cognitive states. Though there have been a variety of approaches proposed for this task, most of these models are trained only for the specific target AUs, and as such they fail to easily adapt to the task of recognition of new AUs (i.e., those not initially used to train the target models). In this paper, we propose a deep learning approach for facial AU detection that can easily and in a fast manner adapt to a new AU or target subject by leveraging only a few labeled samples from the new task (either an AU or subject). To this end, we propose a modeling approach based on the notion of the model-agnostic meta-learning, originally proposed for the general image recognition/detection tasks (e.g., the character recognition from the Omniglot dataset). Specifically, each subject and/or AU is treated as a new learning task and the model learns to adapt based on the knowledge of the previous tasks (the AUs and subjects used to pre-train the target models). Thus, given a new subject or AU, this meta-knowledge (that is shared among training and test tasks) is used to adapt the model to the new task using the notion of deep learning and model-agnostic meta-learning. We show on two benchmark datasets (BP4D and DISFA) for facial AU detection that the proposed approach can be easily adapted to new tasks (AUs/subjects). Using only a few labeled examples from these tasks, the model achieves large improvements over the baselines (i.e., non-adapted models).

Ognjen Rudovic, Meiru Zhang, Bjorn Schuller et al.

Human behavior expression and experience are inherently multi-modal, and characterized by vast individual and contextual heterogeneity. To achieve meaningful human-computer and human-robot interactions, multi-modal models of the users states (e.g., engagement) are therefore needed. Most of the existing works that try to build classifiers for the users states assume that the data to train the models are fully labeled. Nevertheless, data labeling is costly and tedious, and also prone to subjective interpretations by the human coders. This is even more pronounced when the data are multi-modal (e.g., some users are more expressive with their facial expressions, some with their voice). Thus, building models that can accurately estimate the users states during an interaction is challenging. To tackle this, we propose a novel multi-modal active learning (AL) approach that uses the notion of deep reinforcement learning (RL) to find an optimal policy for active selection of the users data, needed to train the target (modality-specific) models. We investigate different strategies for multi-modal data fusion, and show that the proposed model-level fusion coupled with RL outperforms the feature-level and modality-specific models, and the naive AL strategies such as random sampling, and the standard heuristics such as uncertainty sampling. We show the benefits of this approach on the task of engagement estimation from real-world child-robot interactions during an autism therapy. Importantly, we show that the proposed multi-modal AL approach can be used to efficiently personalize the engagement classifiers to the target user using a small amount of actively selected users data.

Ognjen Rudovic, Yuria Utsumi, Ricardo Guerrero et al.

We introduce a novel personalized Gaussian Process Experts (pGPE) model for predicting per-subject ADAS-Cog13 cognitive scores -- a significant predictor of Alzheimer's Disease (AD) in the cognitive domain -- over the future 6, 12, 18, and 24 months. We start by training a population-level model using multi-modal data from previously seen subjects using a base Gaussian Process (GP) regression. Then, we personalize this model by adapting the base GP sequentially over time to a new (target) subject using domain adaptive GPs, and also by training subject-specific GP. While we show that these models achieve improved performance when selectively applied to the forecasting task (one performs better than the other on different subjects/visits), the average performance per model is suboptimal. To this end, we used the notion of meta learning in the proposed pGPE to design a regression-based weighting of these expert models, where the expert weights are optimized for each subject and his/her future visit. The results on a cohort of subjects from the ADNI dataset show that this newly introduced personalized weighting of the expert models leads to large improvements in accurately forecasting future ADAS-Cog13 scores and their fine-grained changes associated with the AD progression. This approach has potential to help identify at-risk patients early and improve the construction of clinical trials for AD.

Behnam Gholami, Pritish Sahu, Ognjen Rudovic et al.

Unsupervised domain adaptation (uDA) models focus on pairwise adaptation settings where there is a single, labeled, source and a single target domain. However, in many real-world settings one seeks to adapt to multiple, but somewhat similar, target domains. Applying pairwise adaptation approaches to this setting may be suboptimal, as they fail to leverage shared information among multiple domains. In this work we propose an information theoretic approach for domain adaptation in the novel context of multiple target domains with unlabeled instances and one source domain with labeled instances. Our model aims to find a shared latent space common to all domains, while simultaneously accounting for the remaining private, domain-specific factors. Disentanglement of shared and private information is accomplished using a unified information-theoretic approach, which also serves to establish a stronger link between the latent representations and the observed data. The resulting model, accompanied by an efficient optimization algorithm, allows simultaneous adaptation from a single source to multiple target domains. We test our approach on three challenging publicly-available datasets, showing that it outperforms several popular domain adaptation methods.

Adria Ruiz, Ognjen Rudovic, Xavier Binefa et al.

We propose a Multi-Instance-Learning (MIL) approach for weakly-supervised learning problems, where a training set is formed by bags (sets of feature vectors or instances) and only labels at bag-level are provided. Specifically, we consider the Multi-Instance Dynamic-Ordinal-Regression (MI-DOR) setting, where the instance labels are naturally represented as ordinal variables and bags are structured as temporal sequences. To this end, we propose Multi-Instance Dynamic Ordinal Random Fields (MI-DORF). In this framework, we treat instance-labels as temporally-dependent latent variables in an Undirected Graphical Model. Different MIL assumptions are modelled via newly introduced high-order potentials relating bag and instance-labels within the energy function of the model. We also extend our framework to address the Partially-Observed MI-DOR problems, where a subset of instance labels are available during training. We show on the tasks of weakly-supervised facial behavior analysis, Facial Action Unit (DISFA dataset) and Pain (UNBC dataset) Intensity estimation, that the proposed framework outperforms alternative learning approaches. Furthermore, we show that MIDORF can be employed to reduce the data annotation efforts in this context by large-scale.

Yuria Utsumi, Ognjen Rudovic, Kelly Peterson et al.

In this paper, we introduce the use of a personalized Gaussian Process model (pGP) to predict per-patient changes in ADAS-Cog13 -- a significant predictor of Alzheimer's Disease (AD) in the cognitive domain -- using data from each patient's previous visits, and testing on future (held-out) data. We start by learning a population-level model using multi-modal data from previously seen patients using a base Gaussian Process (GP) regression. The personalized GP (pGP) is formed by adapting the base GP sequentially over time to a new (target) patient using domain adaptive GPs. We extend this personalized approach to predict the values of ADAS-Cog13 over the future 6, 12, 18, and 24 months. We compare this approach to a GP model trained only on past data of the target patients (tGP), as well as to a new approach that combines pGP with tGP. We find that the new approach, combining pGP with tGP, leads to large improvements in accurately forecasting future ADAS-Cog13 scores.

Eduardo Castelló Ferrer, Ognjen Rudovic, Thomas Hardjono et al.

Robots have potential to revolutionize the way we interact with the world around us. One of their largest potentials is in the domain of mobile health where they can be used to facilitate clinical interventions. However, to accomplish this, robots need to have access to our private data in order to learn from these data and improve their interaction capabilities. Furthermore, to enhance this learning process, the knowledge sharing among multiple robot units is the natural step forward. However, to date, there is no well-established framework which allows for such data sharing while preserving the privacy of the users (e.g., the hospital patients). To this end, we introduce RoboChain - the first learning framework for secure, decentralized and computationally efficient data and model sharing among multiple robot units installed at multiple sites (e.g., hospitals). RoboChain builds upon and combines the latest advances in open data access and blockchain technologies, as well as machine learning. We illustrate this framework using the example of a clinical intervention conducted in a private network of hospitals. Specifically, we lay down the system architecture that allows multiple robot units, conducting the interventions at different hospitals, to perform efficient learning without compromising the data privacy.

Ognjen Rudovic, Jaeryoung Lee, Miles Dai et al.

Robots have great potential to facilitate future therapies for children on the autism spectrum. However, existing robots lack the ability to automatically perceive and respond to human affect, which is necessary for establishing and maintaining engaging interactions. Moreover, their inference challenge is made harder by the fact that many individuals with autism have atypical and unusually diverse styles of expressing their affective-cognitive states. To tackle the heterogeneity in behavioral cues of children with autism, we use the latest advances in deep learning to formulate a personalized machine learning (ML) framework for automatic perception of the childrens affective states and engagement during robot-assisted autism therapy. The key to our approach is a novel shift from the traditional ML paradigm - instead of using 'one-size-fits-all' ML models, our personalized ML framework is optimized for each child by leveraging relevant contextual information (demographics and behavioral assessment scores) and individual characteristics of each child. We designed and evaluated this framework using a dataset of multi-modal audio, video and autonomic physiology data of 35 children with autism (age 3-13) and from 2 cultures (Asia and Europe), participating in a 25-minute child-robot interaction (~500k datapoints). Our experiments confirm the feasibility of the robot perception of affect and engagement, showing clear improvements due to the model personalization. The proposed approach has potential to improve existing therapies for autism by offering more efficient monitoring and summarization of the therapy progress.

Kelly Peterson, Ognjen Rudovic, Ricardo Guerrero et al.

In this paper, we introduce the use of a personalized Gaussian Process model (pGP) to predict the key metrics of Alzheimer's Disease progression (MMSE, ADAS-Cog13, CDRSB and CS) based on each patient's previous visits. We start by learning a population-level model using multi-modal data from previously seen patients using the base Gaussian Process (GP) regression. Then, this model is adapted sequentially over time to a new patient using domain adaptive GPs to form the patient's pGP. We show that this new approach, together with an auto-regressive formulation, leads to significant improvements in forecasting future clinical status and cognitive scores for target patients when compared to modeling the population with traditional GPs.

Daniel Lopez-Martinez, Ognjen Rudovic, Rosalind Picard

Pain is a subjective experience commonly measured through patient's self report. While there exist numerous situations in which automatic pain estimation methods may be preferred, inter-subject variability in physiological and behavioral pain responses has hindered the development of such methods. In this work, we address this problem by introducing a novel personalized multitask machine learning method for pain estimation based on individual physiological and behavioral pain response profiles, and show its advantages in a dataset containing multimodal responses to nociceptive heat pain.

Cuong D. Tran, Ognjen Rudovic, Vladimir Pavlovic

We study the task of unsupervised domain adaptation, where no labeled data from the target domain is provided during training time. To deal with the potential discrepancy between the source and target distributions, both in features and labels, we exploit a copula-based regression framework. The benefits of this approach are two-fold: (a) it allows us to model a broader range of conditional predictive densities beyond the common exponential family, (b) we show how to leverage Sklar's theorem, the essence of the copula formulation relating the joint density to the copula dependency functions, to find effective feature mappings that mitigate the domain mismatch. By transforming the data to a copula domain, we show on a number of benchmark datasets (including human emotion estimation), and using different regression models for prediction, that we can achieve a more robust and accurate estimation of target labels, compared to recently proposed feature transformation (adaptation) methods.

Daniel Lopez Martinez, Ognjen Rudovic, Rosalind Picard

Pain is a personal, subjective experience that is commonly evaluated through visual analog scales (VAS). While this is often convenient and useful, automatic pain detection systems can reduce pain score acquisition efforts in large-scale studies by estimating it directly from the participants' facial expressions. In this paper, we propose a novel two-stage learning approach for VAS estimation: first, our algorithm employs Recurrent Neural Networks (RNNs) to automatically estimate Prkachin and Solomon Pain Intensity (PSPI) levels from face images. The estimated scores are then fed into the personalized Hidden Conditional Random Fields (HCRFs), used to estimate the VAS, provided by each person. Personalization of the model is performed using a newly introduced facial expressiveness score, unique for each person. To the best of our knowledge, this is the first approach to automatically estimate VAS from face images. We show the benefits of the proposed personalized over traditional non-personalized approach on a benchmark dataset for pain analysis from face images.

Dieu Linh Tran, Robert Walecki, Ognjen Rudovic et al.

Human face exhibits an inherent hierarchy in its representations (i.e., holistic facial expressions can be encoded via a set of facial action units (AUs) and their intensity). Variational (deep) auto-encoders (VAE) have shown great results in unsupervised extraction of hierarchical latent representations from large amounts of image data, while being robust to noise and other undesired artifacts. Potentially, this makes VAEs a suitable approach for learning facial features for AU intensity estimation. Yet, most existing VAE-based methods apply classifiers learned separately from the encoded features. By contrast, the non-parametric (probabilistic) approaches, such as Gaussian Processes (GPs), typically outperform their parametric counterparts, but cannot deal easily with large amounts of data. To this end, we propose a novel VAE semi-parametric modeling framework, named DeepCoder, which combines the modeling power of parametric (convolutional) and nonparametric (ordinal GPs) VAEs, for joint learning of (1) latent representations at multiple levels in a task hierarchy1, and (2) classification of multiple ordinal outputs. We show on benchmark datasets for AU intensity estimation that the proposed DeepCoder outperforms the state-of-the-art approaches, and related VAEs and deep learning models.

Adria Ruiz, Ognjen Rudovic, Xavier Binefa et al.

In this paper, we address the Multi-Instance-Learning (MIL) problem when bag labels are naturally represented as ordinal variables (Multi--Instance--Ordinal Regression). Moreover, we consider the case where bags are temporal sequences of ordinal instances. To model this, we propose the novel Multi-Instance Dynamic Ordinal Random Fields (MI-DORF). In this model, we treat instance-labels inside the bag as latent ordinal states. The MIL assumption is modelled by incorporating a high-order cardinality potential relating bag and instance-labels,into the energy function. We show the benefits of the proposed approach on the task of weakly-supervised pain intensity estimation from the UNBC Shoulder-Pain Database. In our experiments, the proposed approach significantly outperforms alternative non-ordinal methods that either ignore the MIL assumption, or do not model dynamic information in target data.

Stefanos Eleftheriadis, Ognjen Rudovic, Marc P. Deisenroth et al.

We address the task of simultaneous feature fusion and modeling of discrete ordinal outputs. We propose a novel Gaussian process(GP) auto-encoder modeling approach. In particular, we introduce GP encoders to project multiple observed features onto a latent space, while GP decoders are responsible for reconstructing the original features. Inference is performed in a novel variational framework, where the recovered latent representations are further constrained by the ordinal output labels. In this way, we seamlessly integrate the ordinal structure in the learned manifold, while attaining robust fusion of the input features. We demonstrate the representation abilities of our model on benchmark datasets from machine learning and affect analysis. We further evaluate the model on the tasks of feature fusion and joint ordinal prediction of facial action units. Our experiments demonstrate the benefits of the proposed approach compared to the state of the art.

Stefanos Eleftheriadis, Ognjen Rudovic, Marc P. Deisenroth et al.

We present a novel approach for supervised domain adaptation that is based upon the probabilistic framework of Gaussian processes (GPs). Specifically, we introduce domain-specific GPs as local experts for facial expression classification from face images. The adaptation of the classifier is facilitated in probabilistic fashion by conditioning the target expert on multiple source experts. Furthermore, in contrast to existing adaptation approaches, we also learn a target expert from available target data solely. Then, a single and confident classifier is obtained by combining the predictions from multiple experts based on their confidence. Learning of the model is efficient and requires no retraining/reweighting of the source classifiers. We evaluate the proposed approach on two publicly available datasets for multi-class (MultiPIE) and multi-label (DISFA) facial expression classification. To this end, we perform adaptation of two contextual factors: 'where' (view) and 'who' (subject). We show in our experiments that the proposed approach consistently outperforms both source and target classifiers, while using as few as 30 target examples. It also outperforms the state-of-the-art approaches for supervised domain adaptation.

Robert Walecki, Ognjen Rudovic, Vladimir Pavlovic et al.

Automated recognition of facial expressions of emotions, and detection of facial action units (AUs), from videos depends critically on modeling of their dynamics. These dynamics are characterized by changes in temporal phases (onset-apex-offset) and intensity of emotion expressions and AUs, the appearance of which may vary considerably among target subjects, making the recognition/detection task very challenging. The state-of-the-art Latent Conditional Random Fields (L-CRF) framework allows one to efficiently encode these dynamics through the latent states accounting for the temporal consistency in emotion expression and ordinal relationships between its intensity levels, these latent states are typically assumed to be either unordered (nominal) or fully ordered (ordinal). Yet, such an approach is often too restrictive. For instance, in the case of AU detection, the goal is to discriminate between the segments of an image sequence in which this AU is active or inactive. While the sequence segments containing activation of the target AU may better be described using ordinal latent states, the inactive segments better be described using unordered (nominal) latent states, as no assumption can be made about their underlying structure (since they can contain either neutral faces or activations of non-target AUs). To address this, we propose the variable-state L-CRF (VSL-CRF) model that automatically selects the optimal latent states for the target image sequence. To reduce the model overfitting either the nominal or ordinal latent states, we propose a novel graph-Laplacian regularization of the latent states. Our experiments on three public expression databases show that the proposed model achieves better generalization performance compared to traditional L-CRFs and other related state-of-the-art models.

Ognjen Rudovic, Maja Pantic, Vladimir Pavlovic

We propose a novel method for automatic pain intensity estimation from facial images based on the framework of kernel Conditional Ordinal Random Fields (KCORF). We extend this framework to account for heteroscedasticity on the output labels(i.e., pain intensity scores) and introduce a novel dynamic features, dynamic ranks, that impose temporal ordinal constraints on the static ranks (i.e., intensity scores). Our experimental results show that the proposed approach outperforms state-of-the art methods for sequence classification with ordinal data and other ordinal regression models. The approach performs significantly better than other models in terms of Intra-Class Correlation measure, which is the most accepted evaluation measure in the tasks of facial behaviour intensity estimation.