One of the most classical results in high-dimensional learning theory provides a closed-form expression for the generalisation error of binary classification with the single-layer teacher-student perceptron on i.i.d. Gaussian inputs. Both Bayes-optimal estimation and empirical risk minimisation (ERM) were extensively analysed for this setting. At the same time, a considerable part of modern machine learning practice concerns multi-class classification. Yet, an analogous analysis for the corresponding multi-class teacher-student perceptron was missing. In this manuscript we fill this gap by deriving and evaluating asymptotic expressions for both the Bayes-optimal and ERM generalisation errors in the high-dimensional regime. For Gaussian teacher weights, we investigate the performance of ERM with both cross-entropy and square losses, and explore the role of ridge regularisation in approaching Bayes-optimality. In particular, we observe that regularised cross-entropy minimisation yields close-to-optimal accuracy. Instead, for a binary teacher we show that a first-order phase transition arises in the Bayes-optimal performance.
Face frontalization consists of synthesizing a frontally-viewed face from an arbitrarily-viewed one. The main contribution of this paper is a frontalization methodology that preserves non-rigid facial deformations in order to boost the performance of visually assisted speech communication. The method alternates between the estimation of (i)~the rigid transformation (scale, rotation, and translation) and (ii)~the non-rigid deformation between an arbitrarily-viewed face and a face model. The method has two important merits: it can deal with non-Gaussian errors in the data and it incorporates a dynamical face deformation model. For that purpose, we use the generalized Student t-distribution in combination with a linear dynamic system in order to account for both rigid head motions and time-varying facial deformations caused by speech production. We propose to use the zero-mean normalized cross-correlation (ZNCC) score to evaluate the ability of the method to preserve facial expressions. The method is thoroughly evaluated and compared with several state of the art methods, either based on traditional geometric models or on deep learning. Moreover, we show that the method, when incorporated into deep learning pipelines, namely lip reading and speech enhancement, improves word recognition and speech intelligibilty scores by a considerable margin. Supplemental material is accessible at https://team.inria.fr/robotlearn/research/facefrontalization-benchmark/
The accuracy of prosodic structure prediction is crucial to the naturalness of synthesized speech in Mandarin text-to-speech system, but now is limited by widely-used sequence-to-sequence framework and error accumulation from previous word segmentation results. In this paper, we propose a span-based Mandarin prosodic structure prediction model to obtain an optimal prosodic structure tree, which can be converted to corresponding prosodic label sequence. Instead of the prerequisite for word segmentation, rich linguistic features are provided by Chinese character-level BERT and sent to encoder with self-attention architecture. On top of this, span representation and label scoring are used to describe all possible prosodic structure trees, of which each tree has its corresponding score. To find the optimal tree with the highest score for a given sentence, a bottom-up CKY-style algorithm is further used. The proposed method can predict prosodic labels of different levels at the same time and accomplish the process directly from Chinese characters in an end-to-end manner. Experiment results on two real-world datasets demonstrate the excellent performance of our span-based method over all sequence-to-sequence baseline approaches.
Some exciting new approaches to neural architectures for the analysis of conversation have been introduced over the past couple of years. These include neural architectures for detecting emotion, dialogue acts, and sentiment polarity. They take advantage of some of the key attributes of contemporary machine learning, such as recurrent neural networks with attention mechanisms and transformer-based approaches. However, while the architectures themselves are extremely promising, the phenomena they have been applied to to date are but a small part of what makes conversation engaging. In this paper we survey these neural architectures and what they have been applied to. On the basis of the social science literature, we then describe what we believe to be the most fundamental and definitional feature of conversation, which is its co-construction over time by two or more interlocutors. We discuss how neural architectures of the sort surveyed could profitably be applied to these more fundamental aspects of conversation, and what this buys us in terms of a better analysis of conversation and even, in the longer term, a better way of generating conversation for a conversational system.
Edges are image locations where the gray value intensity changes suddenly. They are among the most important features to understand and segment an image. Edge detection is a standard task in digital image processing, solved for example using filtering techniques. However, the amount of data to be processed grows rapidly and pushes even supercomputers to their limits. Quantum computing promises exponentially lower memory usage in terms of the number of qubits compared to the number of classical bits. In this paper, we propose a hybrid method for quantum edge detection based on the idea of a quantum artificial neuron. Our method can be practically implemented on quantum computers, especially on those of the current noisy intermediate-scale quantum era. We compare six variants of the method to reduce the number of circuits and thus the time required for the quantum edge detection. Taking advantage of the scalability of our method, we can practically detect edges in images considerably larger than reached before.
Recent advances in domain adaptation establish that requiring a low risk on the source domain and equal feature marginals degrade the adaptation's performance. At the same time, empirical evidence shows that incorporating an unsupervised target domain term that pushes decision boundaries away from the high-density regions, along with relaxed alignment, improves adaptation. In this paper, we theoretically justify such observations via a new bound on the target risk, and we connect two notions of relaxation for divergence, namely $\beta-$relaxed divergences and localization. This connection allows us to incorporate the source domain's categorical structure into the relaxation of the considered divergence, provably resulting in a better handling of the label shift case in particular.
Businesses need content. In various forms and formats and for varied purposes. In fact, the content marketing industry is set to be worth $412.88 billion by the end of 2021. However, according to the Content Marketing Institute, creating engaging content is the #1 challenge that marketers face today. We under-stand that producing great content requires great writers who understand the business and can weave their message into reader (and search engine) friendly content. In this project, the team has attempted to bridge the gap between writers and projects by using AI and ML tools. We used NLP techniques to analyze thou-sands of publicly available business articles (corpora) to extract various defining factors for each writing sample. Through this project we aim to automate the highly time-consuming, and often biased task of manually shortlisting the most suitable writer for a given content writing requirement. We believe that a tool like this will have far reaching positive implications for both parties - businesses looking for suitable talent for niche writing jobs as well as experienced writers and Subject Matter Experts (SMEs) wanting to lend their services to content marketing projects. The business gets the content they need, the content writer/ SME gets a chance to leverage his or her talent, while the reader gets authentic content that adds real value.
Highly accurate different horizon-based wind speed forecasting facilitates a better modern power system. This paper proposed a novel astute hybrid wind speed forecasting model and applied it to different horizons. The proposed hybrid forecasting model decomposes the original wind speed data into IMFs (Intrinsic Mode Function) using Improved Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (ICEEMDAN). We fed the obtained subseries from ICEEMDAN to the transformer network. Each transformer network computes the forecast subseries and then passes to the fusion phase. Get the primary wind speed forecasting from the fusion of individual transformer network forecast subseries. Estimate the residual error values and predict errors using a multilayer perceptron neural network. The forecast error is added to the primary forecast wind speed to leverage the high accuracy of wind speed forecasting. Comparative analysis with real-time Kethanur, India wind farm dataset results reveals the proposed ICEEMDAN-TNF-MLPN-RECS hybrid model's superior performance with MAE=1.7096*10^-07, MAPE=2.8416*10^-06, MRE=2.8416*10^-08, MSE=5.0206*10^-14, and RMSE=2.2407*10^-07 for case study 1 and MAE=6.1565*10^-07, MAPE=9.5005*10^-06, MRE=9.5005*10^-08, MSE=8.9289*10^-13, and RMSE=9.4493*10^-07 for case study 2 enriched wind speed forecasting than state-of-the-art methods and reduces the burden on the power system engineer.
Iterative distributed optimization algorithms involve multiple agents that communicate with each other, over time, in order to minimize/maximize a global objective. In the presence of unreliable communication networks, the Age-of-Information (AoI), which measures the freshness of data received, may be large and hence hinder algorithmic convergence. In this paper, we study the convergence of general distributed gradient-based optimization algorithms in the presence of communication that neither happens periodically nor at stochastically independent points in time. We show that convergence is guaranteed provided the random variables associated with the AoI processes are stochastically dominated by a random variable with finite first moment. This improves on previous requirements of boundedness of more than the first moment. We then introduce stochastically strongly connected (SSC) networks, a new stochastic form of strong connectedness for time-varying networks. We show: If for any $p \ge0$ the processes that describe the success of communication between agents in a SSC network are $\alpha$-mixing with $n^{p-1}\alpha(n)$ summable, then the associated AoI processes are stochastically dominated by a random variable with finite $p$-th moment. In combination with our first contribution, this implies that distributed stochastic gradient descend converges in the presence of AoI, if $\alpha(n)$ is summable.
Engagement is an essential indicator of the Quality-of-Learning Experience (QoLE) and plays a major role in developing intelligent educational interfaces. The number of people learning through Massively Open Online Courses (MOOCs) and other online resources has been increasing rapidly because they provide us with the flexibility to learn from anywhere at any time. This provides a good learning experience for the students. However, such learning interface requires the ability to recognize the level of engagement of the students for a holistic learning experience. This is useful for both students and educators alike. However, understanding engagement is a challenging task, because of its subjectivity and ability to collect data. In this paper, we propose a variety of models that have been trained on an open-source dataset of video screengrabs. Our non-deep learning models are based on the combination of popular algorithms such as Histogram of Oriented Gradient (HOG), Support Vector Machine (SVM), Scale Invariant Feature Transform (SIFT) and Speeded Up Robust Features (SURF). The deep learning methods include Densely Connected Convolutional Networks (DenseNet-121), Residual Network (ResNet-18) and MobileNetV1. We show the performance of each models using a variety of metrics such as the Gini Index, Adjusted F-Measure (AGF), and Area Under receiver operating characteristic Curve (AUC). We use various dimensionality reduction techniques such as Principal Component Analysis (PCA) and t-Distributed Stochastic Neighbor Embedding (t-SNE) to understand the distribution of data in the feature sub-space. Our work will thereby assist the educators and students in obtaining a fruitful and efficient online learning experience.