已知神经模型被过度参数化，最近的工作表明，稀疏的文本到语音（TTS）模型可以超过密集的模型。尽管已经为其他域提出了大量稀疏方法，但这种方法很少在TTS中应用。在这项工作中，我们试图回答以下问题：所选稀疏技术在性能和模型复杂性上的特征是什么？我们比较了Tacotron2基线和应用五种技术的结果。然后，我们通过自然性，清晰度和韵律来评估表现，同时报告模型规模和训练时间。与先前的研究相辅相成，我们发现在训练之前或期间进行修剪可以实现与训练后的修剪相似的性能，并且可以更快地进行培训，同时除去整个神经元降低了性能远不止于删除参数。据我们所知，这是比较文本到语音综合中稀疏范式的第一部作品。

会话问题生成（CQG）是机器通过对话等人类（例如交互式阅读理解）的重要任务。与传统的单转交问题（SQG）相比，CQG更具挑战性的意义，即生成的问题不仅需要有意义，而且要与发生的对话历史保持一致。虽然先前的研究主要集中于如何建模对话的流量和对齐，但迄今为止，尚无对模型必需部分和历史的部分进行全面的研究。我们认为，缩短上下文和历史是至关重要的，因为它可以帮助该模型对对话的一致性进行更多优化。为此，我们提出了一个两阶段CQG框架COHS-CQG，该框架采用COHS模块来缩短输入的上下文和历史记录。特别是，COHS选择连续的句子，并根据其相关性得分通过顶级P策略转弯。我们的模型在答案感和答案环境中都可以在COQA上实现最先进的表演。

对话状态跟踪器是为了跟踪对话中用户目标的设计，是对话系统中的重要组成部分。但是，对话状态跟踪的研究在很大程度上仅限于单形式，其中插槽和老虎机值受知识领域（例如带有餐厅名称和价格范围插槽的餐厅域）的限制，并且由特定的数据库架构定义。在本文中，我们建议将对话状态跟踪的定义扩展到多模式。具体来说，我们介绍了一项新颖的对话状态跟踪任务，以跟踪视频接地对话中提到的视觉对象的信息。每个新的对话说法都可能引入一个新的视频段，新的视觉对象或新对象属性，并且需要一个状态跟踪器来相应地更新这些信息插槽。我们创建了一个新的合成基准测试，并为此任务设计了一个新颖的基线视频 - 底盘变压器网络（VDTN）。 VDTN结合了对象级功能和段级功能，并学习视频和对话之间的上下文依赖性，以生成多模式对话状态。我们为国家生成任务以及一个自我监督的视频理解任务优化了VDTN，该任务恢复了视频段或对象表示。最后，我们培训了VDTN在响应预测任务中使用解码状态。加上全面的消融和定性分析，我们发现了一些有趣的见解，以建立更有能力的多模式对话系统。

神经模块网络（NMN）在图像接地任务中取得了成功，例如在合成图像上的视觉询问（VQA）。但是，在视频接地的对话任务中已经研究了NMN的非常有限的工作。这些任务通过附加的视觉时间差异和语言交叉转移依赖性扩展了传统视觉任务的复杂性。在最新的NMN方法上，我们介绍了视频接地的神经模块网络（VGNMN），以模拟视频基础语言任务中的信息检索过程，作为神经模块的管道。 VGNMN首先分解对话中的所有语言组件，以明确解决任何实体参考并从问题中检测相应的基于动作的输入。检测到的实体和动作被用作实例化神经模块网络并从视频中提取视觉提示的参数。我们的实验表明，VGNMN可以在充满挑战的视频对话基准以及视频质量质量标准测试中实现有希望的表现。

Compared to traditional visual question answering, video-grounded dialogues require additional reasoning over dialogue context to answer questions in a multi-turn setting. Previous approaches to video-grounded dialogues mostly use dialogue context as a simple text input without modelling the inherent information flows at the turn level. In this paper, we propose a novel framework of Reasoning Paths in Dialogue Context (PDC). PDC model discovers information flows among dialogue turns through a semantic graph constructed based on lexical components in each question and answer. PDC model then learns to predict reasoning paths over this semantic graph. Our path prediction model predicts a path from the current turn through past dialogue turns that contain additional visual cues to answer the current question. Our reasoning model sequentially processes both visual and textual information through this reasoning path and the propagated features are used to generate the answer. Our experimental results demonstrate the effectiveness of our method and provide additional insights on how models use semantic dependencies in a dialogue context to retrieve visual cues.

An important class of techniques for resonant anomaly detection in high energy physics builds models that can distinguish between reference and target datasets, where only the latter has appreciable signal. Such techniques, including Classification Without Labels (CWoLa) and Simulation Assisted Likelihood-free Anomaly Detection (SALAD) rely on a single reference dataset. They cannot take advantage of commonly-available multiple datasets and thus cannot fully exploit available information. In this work, we propose generalizations of CWoLa and SALAD for settings where multiple reference datasets are available, building on weak supervision techniques. We demonstrate improved performance in a number of settings with realistic and synthetic data. As an added benefit, our generalizations enable us to provide finite-sample guarantees, improving on existing asymptotic analyses.

With the rise of task-specific pre-training objectives, abstractive summarization models like PEGASUS offer appealing zero-shot performance on downstream summarization tasks. However, the performance of such unsupervised models still lags significantly behind their supervised counterparts. Similarly to the supervised setup, we notice a very high variance in quality among summary candidates from these models whereas only one candidate is kept as the summary output. In this paper, we propose to re-rank summary candidates in an unsupervised manner, aiming to close the performance gap between unsupervised and supervised models. Our approach improves the pre-trained unsupervised PEGASUS by 4.37% to 7.27% relative mean ROUGE across four widely-adopted summarization benchmarks, and achieves relative gains of 7.51% (up to 23.73%) averaged over 30 transfer setups.

The number of international benchmarking competitions is steadily increasing in various fields of machine learning (ML) research and practice. So far, however, little is known about the common practice as well as bottlenecks faced by the community in tackling the research questions posed. To shed light on the status quo of algorithm development in the specific field of biomedical imaging analysis, we designed an international survey that was issued to all participants of challenges conducted in conjunction with the IEEE ISBI 2021 and MICCAI 2021 conferences (80 competitions in total). The survey covered participants' expertise and working environments, their chosen strategies, as well as algorithm characteristics. A median of 72% challenge participants took part in the survey. According to our results, knowledge exchange was the primary incentive (70%) for participation, while the reception of prize money played only a minor role (16%). While a median of 80 working hours was spent on method development, a large portion of participants stated that they did not have enough time for method development (32%). 25% perceived the infrastructure to be a bottleneck. Overall, 94% of all solutions were deep learning-based. Of these, 84% were based on standard architectures. 43% of the respondents reported that the data samples (e.g., images) were too large to be processed at once. This was most commonly addressed by patch-based training (69%), downsampling (37%), and solving 3D analysis tasks as a series of 2D tasks. K-fold cross-validation on the training set was performed by only 37% of the participants and only 50% of the participants performed ensembling based on multiple identical models (61%) or heterogeneous models (39%). 48% of the respondents applied postprocessing steps.

Purpose: The aim of this study was to demonstrate the utility of unsupervised domain adaptation (UDA) in automated knee osteoarthritis (OA) phenotype classification using a small dataset (n=50). Materials and Methods: For this retrospective study, we collected 3,166 three-dimensional (3D) double-echo steady-state magnetic resonance (MR) images from the Osteoarthritis Initiative dataset and 50 3D turbo/fast spin-echo MR images from our institute (in 2020 and 2021) as the source and target datasets, respectively. For each patient, the degree of knee OA was initially graded according to the MRI Osteoarthritis Knee Score (MOAKS) before being converted to binary OA phenotype labels. The proposed UDA pipeline included (a) pre-processing, which involved automatic segmentation and region-of-interest cropping; (b) source classifier training, which involved pre-training phenotype classifiers on the source dataset; (c) target encoder adaptation, which involved unsupervised adaption of the source encoder to the target encoder and (d) target classifier validation, which involved statistical analysis of the target classification performance evaluated by the area under the receiver operating characteristic curve (AUROC), sensitivity, specificity and accuracy. Additionally, a classifier was trained without UDA for comparison. Results: The target classifier trained with UDA achieved improved AUROC, sensitivity, specificity and accuracy for both knee OA phenotypes compared with the classifier trained without UDA. Conclusion: The proposed UDA approach improves the performance of automated knee OA phenotype classification for small target datasets by utilising a large, high-quality source dataset for training. The results successfully demonstrated the advantages of the UDA approach in classification on small datasets.

Deep neural networks (DNNs) have rapidly become a \textit{de facto} choice for medical image understanding tasks. However, DNNs are notoriously fragile to the class imbalance in image classification. We further point out that such imbalance fragility can be amplified when it comes to more sophisticated tasks such as pathology localization, as imbalances in such problems can have highly complex and often implicit forms of presence. For example, different pathology can have different sizes or colors (w.r.t.the background), different underlying demographic distributions, and in general different difficulty levels to recognize, even in a meticulously curated balanced distribution of training data. In this paper, we propose to use pruning to automatically and adaptively identify \textit{hard-to-learn} (HTL) training samples, and improve pathology localization by attending them explicitly, during training in \textit{supervised, semi-supervised, and weakly-supervised} settings. Our main inspiration is drawn from the recent finding that deep classification models have difficult-to-memorize samples and those may be effectively exposed through network pruning \cite{hooker2019compressed} - and we extend such observation beyond classification for the first time. We also present an interesting demographic analysis which illustrates HTLs ability to capture complex demographic imbalances. Our extensive experiments on the Skin Lesion Localization task in multiple training settings by paying additional attention to HTLs show significant improvement of localization performance by $\sim$2-3\%.