Utterance-level neural confidence measure for end-to-end children speech recognition

Confidence measure is a performance index of particular importance for automatic speech recognition (ASR) systems deployed in real-world scenarios. In the present study, utterance-level neural confidence measure (NCM) in end-to-end automatic speech recognition (E2E ASR) is investigated. The E2E system adopts the joint CTC-attention Transformer architecture. The prediction of NCM is formulated as a task of binary classification, i.e., accept/reject the input utterance, based on a set of predictor features acquired during the ASR decoding process. The investigation is focused on evaluating and comparing the efficacies of predictor features that are derived from different internal and external modules of the E2E system. Experiments are carried out on children speech, for which state-of-the-art ASR systems show less than satisfactory performance and robust confidence measure is particularly useful. It is noted that predictor features related to acoustic information of speech play a more important role in estimating confidence measure than those related to linguistic information. N-best score features show significantly better performance than single-best ones. It has also been shown that the metrics of EER and AUC are not appropriate to evaluate the NCM of a mismatched ASR with significant performance gap.

Stellar loci IV. Red giant stars

In the fourth paper of this series, we present the metallicity-dependent Sloan Digital Sky Survey (SDSS) stellar color loci of red giant stars, using a spectroscopic sample of red giants in the SDSS Stripe 82 region. The stars span a range of 0.55 -- 1.2 mag in color g-i, -0.3 -- -2.5 in metallicity [Fe/H], and have values of surface gravity log g smaller than 3.5 dex. As in the case of main-sequence (MS) stars, the intrinsic widths of loci of red giants are also found to be quite narrow, a few mmag at maximum. There are however systematic differences between the metallicity-dependent stellar loci of red giants and MS stars. The colors of red giants are less sensitive to metallicity than those of MS stars. With good photometry, photometric metallicities of red giants can be reliably determined by fitting the u-g, g-r, r-i, and i-z colors simultaneously to an accuracy of 0.2 -- 0.25 dex, comparable to the precision achievable with low-resolution spectroscopy for a signal-to-noise ratio of 10. By comparing fitting results to the stellar loci of red giants and MS stars, we propose a new technique to discriminate between red giants and MS stars based on the SDSS photometry. The technique achieves completeness of ~ 70 per cent and efficiency of ~ 80 per cent in selecting metal-poor red giant stars of [Fe/H] $le$ -1.2. It thus provides an important tool to probe the structure and assemblage history of the Galactic halo using red giant stars.

Nonmonotone Local Minimax Methods for Finding Multiple Saddle Points

In this paper, combining normalized nonmonotone search strategies with the Barzilai--Borwein-type step-size, a novel local minimax method (LMM), which is globally convergent, is proposed to find multiple (unstable) saddle points of nonconvex functionals in Hilbert spaces. Compared to traditional LMMs, this approach does not require the strict decrease of the objective functional value at each iterative step. Firstly, by introducing two kinds of normalized nonmonotone step-size search strategies to replace normalized monotone decrease conditions adopted in traditional LMMs, two types of nonmonotone LMMs are constructed. Their feasibility and convergence results are rigorously carried out. Secondly, in order to speed up the convergence of the nonmonotone LMMs, a globally convergent Barzilai--Borwein-type LMM (GBBLMM) is presented by explicitly constructing the Barzilai--Borwein-type step-size as a trial step-size of the normalized nonmonotone step-size search strategy in each iteration. Finally, the GBBLMM is implemented to find multiple unstable solutions of two classes of semilinear elliptic boundary value problems with variational structures: one is the semilinear elliptic equations with the homogeneous Dirichlet boundary condition and another is the linear elliptic equations with semilinear Neumann boundary conditions. Extensive numerical results indicate that our approach is very effective and speeds up the LMM significantly.

Cherenkov Gravitational Radiation During the Radiation Era

Cherenkov radiation may occur whenever the source is moving faster than the waves it generates. In a radiation dominated universe, with equation-of-state $w = 1/3$, we have recently shown that the Bardeen scalar-metric perturbations contribute to the linearized Weyl tensor in such a manner that its wavefront propagates at acoustic speed $sqrt{w}=1/sqrt{3}$. In this work, we explicitly compute the shape of the Bardeen Cherenkov cone and wedge generated respectively by a supersonic point mass (approximating a primordial black hole) and a straight Nambu-Goto wire (approximating a cosmic string) moving perpendicular to its length. When the black hole or cosmic string is moving at ultra-relativistic speeds, we also calculate explicitly the sudden surge of scalar-metric induced tidal forces on a pair of test particles due to the passing Cherenkov shock wave. These forces can stretch or compress, depending on the orientation of the masses relative to the shock fronts normal.

DSKReG: Differentiable Sampling on Knowledge Graph for Recommendation with Relational GNN

In the information explosion era, recommender systems (RSs) are widely studied and applied to discover user-preferred information. A RS performs poorly when suffering from the cold-start issue, which can be alleviated if incorporating Knowledge Graphs (KGs) as side information. However, most existing works neglect the facts that node degrees in KGs are skewed and massive amount of interactions in KGs are recommendation-irrelevant. To address these problems, in this paper, we propose Differentiable Sampling on Knowledge Graph for Recommendation with Relational GNN (DSKReG) that learns the relevance distribution of connected items from KGs and samples suitable items for recommendation following this distribution. We devise a differentiable sampling strategy, which enables the selection of relevant items to be jointly optimized with the model training procedure. The experimental results demonstrate that our model outperforms state-of-the-art KG-based recommender systems. The code is available online at https://github.com/YuWang-1024/DSKReG.

Fast and Flexible Human Pose Estimation with HyperPose

Estimating human pose is an important yet challenging task in multimedia applications. Existing pose estimation libraries target reproducing standard pose estimation algorithms. When it comes to customising these algorithms for real-world applications, none of the existing libraries can offer both the flexibility of developing custom pose estimation algorithms and the high-performance of executing these algorithms on commodity devices. In this paper, we introduce Hyperpose, a novel flexible and high-performance pose estimation library. Hyperpose provides expressive Python APIs that enable developers to easily customise pose estimation algorithms for their applications. It further provides a model inference engine highly optimised for real-time pose estimation. This engine can dynamically dispatch carefully designed pose estimation tasks to CPUs and GPUs, thus automatically achieving high utilisation of hardware resources irrespective of deployment environments. Extensive evaluation results show that Hyperpose can achieve up to 3.1x~7.3x higher pose estimation throughput compared to state-of-the-art pose estimation libraries without compromising estimation accuracy. By 2021, Hyperpose has received over 1000 stars on GitHub and attracted users from both industry and academy.

Effects of the Crystalline Electric Field in the $KErTe_{2}$ Quantum Spin Liquid Candidate

In this paper, we performed thermodynamic and electron spin resonance (ESR) measurements to study low-energy magnetic excitations, which were significantly affected by crystalline electric field (CEF) excitations due to relatively small gaps between the CEF ground state and the excited states. Based on the CEF and mean-field (MF) theories, we analyzed systematically and consistently the ESR experiments and thermodynamic measurements including susceptibility, magnetization, and heat capacity. The CEF parameters were successfully extracted by fitting high-temperature (> 20 K) susceptibilities in the ab-plane and along the c-axis, allowing to determine the Lande factors ($g_{ab,calc}$ = 5.98(7) and $g_{c,calc}$ = 2.73(3)). These values were consistent with the values of Lande factors determined by ESR experiments ($g_{ab,exp}$ = 5.69 and $g_{c,exp}$ = 2.75). By applying the CEF and MF theories to the susceptibility and magnetization results, we estimated the anisotropic spin-exchange energies and found that the CEF excitations in ce{KErTe2} played a decisive role in the magnetism above 3 K, while the low-temperature magnetism below 10 K was gradually correlated with the anisotropic spin-exchange interactions. The CEF excitations were demonstrated in the low-temperature heat capacity, where both the positions of two broad peaks and their magnetic field dependence well corroborated our calculations. The present study provides a basis to explore the enriched magnetic and electronic properties of the QSL family.

Deep Natural Language Processing for LinkedIn Search

Many search systems work with large amounts of natural language data, e.g., search queries, user profiles, and documents. Building a successful search system requires a thorough understanding of textual data semantics, where deep learning based natural language processing techniques (deep NLP) can be of great help. In this paper, we introduce a comprehensive study for applying deep NLP techniques to five representative tasks in search systems: query intent prediction (classification), query tagging (sequential tagging), document ranking (ranking), query auto completion (language modeling), and query suggestion (sequence to sequence). We also introduce BERT pre-training as a sixth task that can be applied to many of the other tasks. Through the model design and experiments of the six tasks, readers can find answers to four important questions: (1). When is deep NLP helpful/not helpful in search systems? (2). How to address latency challenges? (3). How to ensure model robustness? This work builds on existing efforts of LinkedIn search, and is tested at scale on LinkedIns commercial search engines. We believe our experiences can provide useful insights for the industry and research communities.

Continuous-Time Sequential Recommendation with Temporal Graph Collaborative Transformer

In order to model the evolution of user preference, we should learn user/item embeddings based on time-ordered item purchasing sequences, which is defined as Sequential Recommendation (SR) problem. Existing methods leverage sequential patterns to model item transitions. However, most of them ignore crucial temporal collaborative signals, which are latent in evolving user-item interactions and coexist with sequential patterns. Therefore, we propose to unify sequential patterns and temporal collaborative signals to improve the quality of recommendation, which is rather challenging. Firstly, it is hard to simultaneously encode sequential patterns and collaborative signals. Secondly, it is non-trivial to express the temporal effects of collaborative signals. Hence, we design a new framework Temporal Graph Sequential Recommender (TGSRec) upon our defined continuous-time bi-partite graph. We propose a novel Temporal Collaborative Trans-former (TCT) layer in TGSRec, which advances the self-attention mechanism by adopting a novel collaborative attention. TCT layer can simultaneously capture collaborative signals from both users and items, as well as considering temporal dynamics inside sequential patterns. We propagate the information learned fromTCTlayerover the temporal graph to unify sequential patterns and temporal collaborative signals. Empirical results on five datasets show that TGSRec significantly outperforms other baselines, in average up to 22.5% and 22.1%absolute improvements in Recall@10and MRR, respectively.

Contrastive Self-supervised Sequential Recommendation with Robust Augmentation

Sequential Recommendationdescribes a set of techniques to model dynamic user behavior in order to predict future interactions in sequential user data. At their core, such approaches model transition probabilities between items in a sequence, whether through Markov chains, recurrent networks, or more recently, Transformers. However both old and new issues remain, including data-sparsity and noisy data; such issues can impair the performance, especially in complex, parameter-hungry models. In this paper, we investigate the application of contrastive Self-Supervised Learning (SSL) to the sequential recommendation, as a way to alleviate some of these issues. Contrastive SSL constructs augmentations from unlabelled instances, where agreements among positive pairs are maximized. It is challenging to devise a contrastive SSL framework for a sequential recommendation, due to its discrete nature, correlations among items, and skewness of length distributions. To this end, we propose a novel framework, Contrastive Self-supervised Learning for sequential Recommendation (CoSeRec). We introduce two informative augmentation operators leveraging item correlations to create high-quality views for contrastive learning. Experimental results on three real-world datasets demonstrate the effectiveness of the proposed method on improving model performance and the robustness against sparse and noisy data. Our implementation is available online at url{https://github.com/YChen1993/CoSeRec}