测量自动语音识别（ASR）系统质量对于创建用户满意的语音驱动应用程序至关重要。传统上，单词错误率（WER）用于评估ASR系统质量；但是，它有时与用户对转录质量的看法/判断息息相关。这是因为wer平等地称重每个单词，并且不考虑对用户感知产生更高影响的语义正确性。在这项工作中，我们提出评估ASR输出的质量，可以通过使用参考的语义向量与从预训练的语言模型中提取的假设之间的距离来测量语义正确性。我们对71K和36K用户注释的ASR输出质量的实验结果表明，与WER相比，Semdist与用户感知的相关性更高。我们还表明，与WER相比，Semdist与下游自然语言理解（NLU）任务具有更高的相关性。

Reinforcement learning (RL) problems can be challenging without well-shaped rewards. Prior work on provably efficient RL methods generally proposes to address this issue with dedicated exploration strategies. However, another way to tackle this challenge is to reformulate it as a multi-task RL problem, where the task space contains not only the challenging task of interest but also easier tasks that implicitly function as a curriculum. Such a reformulation opens up the possibility of running existing multi-task RL methods as a more efficient alternative to solving a single challenging task from scratch. In this work, we provide a theoretical framework that reformulates a single-task RL problem as a multi-task RL problem defined by a curriculum. Under mild regularity conditions on the curriculum, we show that sequentially solving each task in the multi-task RL problem is more computationally efficient than solving the original single-task problem, without any explicit exploration bonuses or other exploration strategies. We also show that our theoretical insights can be translated into an effective practical learning algorithm that can accelerate curriculum learning on simulated robotic tasks.

One of the key challenges in deploying RL to real-world applications is to adapt to variations of unknown environment contexts, such as changing terrains in robotic tasks and fluctuated bandwidth in congestion control. Existing works on adaptation to unknown environment contexts either assume the contexts are the same for the whole episode or assume the context variables are Markovian. However, in many real-world applications, the environment context usually stays stable for a stochastic period and then changes in an abrupt and unpredictable manner within an episode, resulting in a segment structure, which existing works fail to address. To leverage the segment structure of piecewise stable context in real-world applications, in this paper, we propose a \textit{\textbf{Se}gmented \textbf{C}ontext \textbf{B}elief \textbf{A}ugmented \textbf{D}eep~(SeCBAD)} RL method. Our method can jointly infer the belief distribution over latent context with the posterior over segment length and perform more accurate belief context inference with observed data within the current context segment. The inferred belief context can be leveraged to augment the state, leading to a policy that can adapt to abrupt variations in context. We demonstrate empirically that SeCBAD can infer context segment length accurately and outperform existing methods on a toy grid world environment and Mujuco tasks with piecewise-stable context.

In recent years, large amounts of effort have been put into pushing forward the real-world application of dynamic digital human (DDH). However, most current quality assessment research focuses on evaluating static 3D models and usually ignores motion distortions. Therefore, in this paper, we construct a large-scale dynamic digital human quality assessment (DDH-QA) database with diverse motion content as well as multiple distortions to comprehensively study the perceptual quality of DDHs. Both model-based distortion (noise, compression) and motion-based distortion (binding error, motion unnaturalness) are taken into consideration. Ten types of common motion are employed to drive the DDHs and a total of 800 DDHs are generated in the end. Afterward, we render the video sequences of the distorted DDHs as the evaluation media and carry out a well-controlled subjective experiment. Then a benchmark experiment is conducted with the state-of-the-art video quality assessment (VQA) methods and the experimental results show that existing VQA methods are limited in assessing the perceptual loss of DDHs. The database will be made publicly available to facilitate future research.

Conversational recommender systems (CRSs) often utilize external knowledge graphs (KGs) to introduce rich semantic information and recommend relevant items through natural language dialogues. However, original KGs employed in existing CRSs are often incomplete and sparse, which limits the reasoning capability in recommendation. Moreover, only few of existing studies exploit the dialogue context to dynamically refine knowledge from KGs for better recommendation. To address the above issues, we propose the Variational Reasoning over Incomplete KGs Conversational Recommender (VRICR). Our key idea is to incorporate the large dialogue corpus naturally accompanied with CRSs to enhance the incomplete KGs; and perform dynamic knowledge reasoning conditioned on the dialogue context. Specifically, we denote the dialogue-specific subgraphs of KGs as latent variables with categorical priors for adaptive knowledge graphs refactor. We propose a variational Bayesian method to approximate posterior distributions over dialogue-specific subgraphs, which not only leverages the dialogue corpus for restructuring missing entity relations but also dynamically selects knowledge based on the dialogue context. Finally, we infuse the dialogue-specific subgraphs to decode the recommendation and responses. We conduct experiments on two benchmark CRSs datasets. Experimental results confirm the effectiveness of our proposed method.

Over the past few years, developing a broad, universal, and general-purpose computer vision system has become a hot topic. A powerful universal system would be capable of solving diverse vision tasks simultaneously without being restricted to a specific problem or a specific data domain, which is of great importance in practical real-world computer vision applications. This study pushes the direction forward by concentrating on the million-scale multi-domain universal object detection problem. The problem is not trivial due to its complicated nature in terms of cross-dataset category label duplication, label conflicts, and the hierarchical taxonomy handling. Moreover, what is the resource-efficient way to utilize emerging large pre-trained vision models for million-scale cross-dataset object detection remains an open challenge. This paper tries to address these challenges by introducing our practices in label handling, hierarchy-aware loss design and resource-efficient model training with a pre-trained large model. Our method is ranked second in the object detection track of Robust Vision Challenge 2022 (RVC 2022). We hope our detailed study would serve as an alternative practice paradigm for similar problems in the community. The code is available at https://github.com/linfeng93/Large-UniDet.

Data-Free Class Incremental Learning (DFCIL) aims to sequentially learn tasks with access only to data from the current one. DFCIL is of interest because it mitigates concerns about privacy and long-term storage of data, while at the same time alleviating the problem of catastrophic forgetting in incremental learning. In this work, we introduce robust saliency guidance for DFCIL and propose a new framework, which we call RObust Saliency Supervision (ROSS), for mitigating the negative effect of saliency drift. Firstly, we use a teacher-student architecture leveraging low-level tasks to supervise the model with global saliency. We also apply boundary-guided saliency to protect it from drifting across object boundaries at intermediate layers. Finally, we introduce a module for injecting and recovering saliency noise to increase robustness of saliency preservation. Our experiments demonstrate that our method can retain better saliency maps across tasks and achieve state-of-the-art results on the CIFAR-100, Tiny-ImageNet and ImageNet-Subset DFCIL benchmarks. Code will be made publicly available.

Vision Transformers convert images to sequences by slicing them into patches. The size of these patches controls a speed/accuracy tradeoff, with smaller patches leading to higher accuracy at greater computational cost, but changing the patch size typically requires retraining the model. In this paper, we demonstrate that simply randomizing the patch size at training time leads to a single set of weights that performs well across a wide range of patch sizes, making it possible to tailor the model to different compute budgets at deployment time. We extensively evaluate the resulting model, which we call FlexiViT, on a wide range of tasks, including classification, image-text retrieval, open-world detection, panoptic segmentation, and semantic segmentation, concluding that it usually matches, and sometimes outperforms, standard ViT models trained at a single patch size in an otherwise identical setup. Hence, FlexiViT training is a simple drop-in improvement for ViT that makes it easy to add compute-adaptive capabilities to most models relying on a ViT backbone architecture. Code and pre-trained models are available at https://github.com/google-research/big_vision

Recognition of facial expression is a challenge when it comes to computer vision. The primary reasons are class imbalance due to data collection and uncertainty due to inherent noise such as fuzzy facial expressions and inconsistent labels. However, current research has focused either on the problem of class imbalance or on the problem of uncertainty, ignoring the intersection of how to address these two problems. Therefore, in this paper, we propose a framework based on Resnet and Attention to solve the above problems. We design weight for each class. Through the penalty mechanism, our model will pay more attention to the learning of small samples during training, and the resulting decrease in model accuracy can be improved by a Convolutional Block Attention Module (CBAM). Meanwhile, our backbone network will also learn an uncertain feature for each sample. By mixing uncertain features between samples, the model can better learn those features that can be used for classification, thus suppressing uncertainty. Experiments show that our method surpasses most basic methods in terms of accuracy on facial expression data sets (e.g., AffectNet, RAF-DB), and it also solves the problem of class imbalance well.

Attention-based arbitrary style transfer studies have shown promising performance in synthesizing vivid local style details. They typically use the all-to-all attention mechanism: each position of content features is fully matched to all positions of style features. However, all-to-all attention tends to generate distorted style patterns and has quadratic complexity. It virtually limits both the effectiveness and efficiency of arbitrary style transfer. In this paper, we rethink what kind of attention mechanism is more appropriate for arbitrary style transfer. Our answer is a novel all-to-key attention mechanism: each position of content features is matched to key positions of style features. Specifically, it integrates two newly proposed attention forms: distributed and progressive attention. Distributed attention assigns attention to multiple key positions; Progressive attention pays attention from coarse to fine. All-to-key attention promotes the matching of diverse and reasonable style patterns and has linear complexity. The resultant module, dubbed StyA2K, has fine properties in rendering reasonable style textures and maintaining consistent local structure. Qualitative and quantitative experiments demonstrate that our method achieves superior results than state-of-the-art approaches.