U2++: Unified Two-pass Bidirectional End-to-end Model for Speech Recognition

The unified streaming and non-streaming two-pass (U2) end-to-end model for speech recognition has shown great performance in terms of streaming capability, accuracy, real-time factor (RTF), and latency. In this paper, we present U2++, an enhanced version of U2 to further improve the accuracy. The core idea of U2++ is to use the forward and the backward information of the labeling sequences at the same time at training to learn richer information, and combine the forward and backward prediction at decoding to give more accurate recognition results. We also proposed a new data augmentation method called SpecSub to help the U2++ model to be more accurate and robust. Our experiments show that, compared with U2, U2++ shows faster convergence at training, better robustness to the decoding method, as well as consistent 5\% - 8\% word error rate reduction gain over U2. On the experiment of AISHELL-1, we achieve a 4.63\% character error rate (CER) with a non-streaming setup and 5.05\% with a streaming setup with 320ms latency by U2++. To the best of our knowledge, 5.05\% is the best-published streaming result on the AISHELL-1 test set.

Near-optimal Representation Learning for Linear Bandits and Linear RL

This paper studies representation learning for multi-task linear bandits and multi-task episodic RL with linear value function approximation. We first consider the setting where we play $M$ linear bandits with dimension $d$ concurrently, and these bandits share a common $k$-dimensional linear representation so that $k\ll d$ and $k \ll M$. We propose a sample-efficient algorithm, MTLR-OFUL, which leverages the shared representation to achieve $\tilde{O}(M\sqrt{dkT} + d\sqrt{kMT} )$ regret, with $T$ being the number of total steps. Our regret significantly improves upon the baseline $\tilde{O}(Md\sqrt{T})$ achieved by solving each task independently. We further develop a lower bound that shows our regret is near-optimal when $d > M$. Furthermore, we extend the algorithm and analysis to multi-task episodic RL with linear value function approximation under low inherent Bellman error \citep{zanette2020learning}. To the best of our knowledge, this is the first theoretical result that characterizes the benefits of multi-task representation learning for exploration in RL with function approximation.

WeNet: Production First and Production Ready End-to-End Speech Recognition Toolkit

In this paper, we present a new open source, production first and production ready end-to-end (E2E) speech recognition toolkit named WeNet. The main motivation of WeNet is to close the gap between the research and the production of E2E speech recognition models. WeNet provides an efficient way to ship ASR applications in several real-world scenarios, which is the main difference and advantage to other open source E2E speech recognition toolkits. This paper introduces WeNet from three aspects, including model architecture, framework design and performance metrics. Our experiments on AISHELL-1 using WeNet, not only give a promising character error rate (CER) on a unified streaming and non-streaming two pass (U2) E2E model but also show reasonable RTF and latency, both of these aspects are favored for production adoption. The toolkit is publicly available at https://github.com/mobvoi/wenet.

Modeling Method for the Coupling Relations of Microgrid Cyber-Physical Systems Driven by Hybrid Spatiotemporal Events

The essence of the microgrid cyber-physical system (CPS) lies in the cyclical conversion of information flow and energy flow. Most of the existing coupling models are modeled with static networks and interface structures, in which the closed-loop data flow characteristic is not fully considered. It is difficult for these models to accurately describe spatiotemporal deduction processes, such as microgrid CPS attack identification, risk propagation, safety assessment, defense control, and cascading failure. To address this problem, a modeling method for the coupling relations of microgrid CPS driven by hybrid spatiotemporal events is proposed in the present work. First, according to the topological correlation and coupling logic of the microgrid CPS, the cyclical conversion mechanism of information flow and energy flow is analyzed, and a microgrid CPS architecture with multi-agents as the core is constructed. Next, the spatiotemporal evolution characteristic of the CPS is described by hybrid automata, and the task coordination mechanism of the multi-agent CPS terminal is designed. On this basis, a discrete-continuous correlation and terminal structure characteristic representation method of the CPS based on heterogeneous multi-groups are then proposed. Finally, four spatiotemporal events, namely state perception, network communication, intelligent decision-making, and action control, are defined. Considering the constraints of the temporal conversion of information flow and energy flow, a microgrid CPS coupling model is established, the effectiveness of which is verified by simulating false data injection attack (FDIA) scenarios.

Compositional Prototype Network with Multi-view Comparision for Few-Shot Point Cloud Semantic Segmentation

Point cloud segmentation is a fundamental visual understanding task in 3D vision. A fully supervised point cloud segmentation network often requires a large amount of data with point-wise annotations, which is expensive to obtain. In this work, we present the Compositional Prototype Network that can undertake point cloud segmentation with only a few labeled training data. Inspired by the few-shot learning literature in images, our network directly transfers label information from the limited training data to unlabeled test data for prediction. The network decomposes the representations of complex point cloud data into a set of local regional representations and utilizes them to calculate the compositional prototypes of a visual concept. Our network includes a key Multi-View Comparison Component that exploits the redundant views of the support set. To evaluate the proposed method, we create a new segmentation benchmark dataset, ScanNet-$6^i$, which is built upon ScanNet dataset. Extensive experiments show that our method outperforms baselines with a significant advantage. Moreover, when we use our network to handle the long-tail problem in a fully supervised point cloud segmentation dataset, it can also effectively boost the performance of the few-shot classes.

Relation Extraction with Contextualized Relation Embedding (CRE)

Relation extraction is the task of identifying relation instance between two entities given a corpus whereas Knowledge base modeling is the task of representing a knowledge base, in terms of relations between entities. This paper proposes an architecture for the relation extraction task that integrates semantic information with knowledge base modeling in a novel manner. Existing approaches for relation extraction either do not utilize knowledge base modelling or use separately trained KB models for the RE task. We present a model architecture that internalizes KB modeling in relation extraction. This model applies a novel approach to encode sentences into contextualized relation embeddings, which can then be used together with parameterized entity embeddings to score relation instances. The proposed CRE model achieves state of the art performance on datasets derived from The New York Times Annotated Corpus and FreeBase. The source code has been made available.

Efficient Reinforcement Learning in Factored MDPs with Application to Constrained RL

Reinforcement learning (RL) in episodic, factored Markov decision processes (FMDPs) is studied. We propose an algorithm called FMDP-BF, which leverages the factorization structure of FMDP. The regret of FMDP-BF is shown to be exponentially smaller than that of optimal algorithms designed for non-factored MDPs, and improves on the best previous result for FMDPs~\citep{osband2014near} by a factored of $\sqrt{H|\mathcal{S}_i|}$, where $|\mathcal{S}_i|$ is the cardinality of the factored state subspace and $H$ is the planning horizon. To show the optimality of our bounds, we also provide a lower bound for FMDP, which indicates that our algorithm is near-optimal w.r.t. timestep $T$, horizon $H$ and factored state-action subspace cardinality. Finally, as an application, we study a new formulation of constrained RL, known as RL with knapsack constraints (RLwK), and provides the first sample-efficient algorithm based on FMDP-BF.

Noisy Agents: Self-supervised Exploration by Predicting Auditory Events

Humans integrate multiple sensory modalities (e.g. visual and audio) to build a causal understanding of the physical world. In this work, we propose a novel type of intrinsic motivation for Reinforcement Learning (RL) that encourages the agent to understand the causal effect of its actions through auditory event prediction. First, we allow the agent to collect a small amount of acoustic data and use K-means to discover underlying auditory event clusters. We then train a neural network to predict the auditory events and use the prediction errors as intrinsic rewards to guide RL exploration. Experimental results on Atari games show that our new intrinsic motivation significantly outperforms several state-of-the-art baselines. We further visualize our noisy agents' behavior in a physics environment and demonstrate that our newly designed intrinsic reward leads to the emergence of physical interaction behaviors (e.g. contact with objects).

(Locally) Differentially Private Combinatorial Semi-Bandits

In this paper, we study Combinatorial Semi-Bandits (CSB) that is an extension of classic Multi-Armed Bandits (MAB) under Differential Privacy (DP) and stronger Local Differential Privacy (LDP) setting. Since the server receives more information from users in CSB, it usually causes additional dependence on the dimension of data, which is a notorious side-effect for privacy preserving learning. However for CSB under two common smoothness assumptions \cite{kveton2015tight,chen2016combinatorial}, we show it is possible to remove this side-effect. In detail, for $B_{\infty}$-bounded smooth CSB under either $\varepsilon$-LDP or $\varepsilon$-DP, we prove the optimal regret bound is $\Theta(\frac{mB^2_{\infty}\ln T } {\Delta\epsilon^2})$ or $\tilde{\Theta}(\frac{mB^2_{\infty}\ln T} { \Delta\epsilon})$ respectively, where $T$ is time period, $\Delta$ is the gap of rewards and $m$ is the number of base arms, by proposing novel algorithms and matching lower bounds. For $B_1$-bounded smooth CSB under $\varepsilon$-DP, we also prove the optimal regret bound is $\tilde{\Theta}(\frac{mKB^2_1\ln T} {\Delta\epsilon})$ with both upper bound and lower bound, where $K$ is the maximum number of feedback in each round. All above results nearly match corresponding non-private optimal rates, which imply there is no additional price for (locally) differentially private CSB in above common settings.