自我监督的方法已通过端到端监督学习的图像分类显着缩小了差距。但是，在人类动作视频的情况下，外观和运动都是变化的重要因素，因此该差距仍然很大。这样做的关键原因之一是，采样对类似的视频剪辑，这是许多自我监督的对比学习方法所需的步骤，目前是保守的，以避免误报。一个典型的假设是，类似剪辑仅在单个视频中暂时关闭，从而导致运动相似性的示例不足。为了减轻这种情况，我们提出了SLIC，这是一种基于聚类的自我监督的对比度学习方法，用于人类动作视频。我们的关键贡献是，我们通过使用迭代聚类来分组类似的视频实例来改善传统的视频内积极采样。这使我们的方法能够利用集群分配中的伪标签来取样更艰难的阳性和负面因素。在UCF101上，SLIC的表现优于最先进的视频检索基线 +15.4％，而直接转移到HMDB51时，SLIC检索基线的率高为15.4％， +5.7％。通过用于动作分类的端到端登录，SLIC在UCF101上获得了83.2％的TOP-1准确性（+0.8％），而HMDB51（+1.6％）上的fric fineTuns in top-1 finetuning。在动力学预处理后，SLIC还与最先进的行动分类竞争。

Differentiable Architecture Search (DARTS) has attracted considerable attention as a gradient-based Neural Architecture Search (NAS) method. Since the introduction of DARTS, there has been little work done on adapting the action space based on state-of-art architecture design principles for CNNs. In this work, we aim to address this gap by incrementally augmenting the DARTS search space with micro-design changes inspired by ConvNeXt and studying the trade-off between accuracy, evaluation layer count, and computational cost. To this end, we introduce the Pseudo-Inverted Bottleneck conv block intending to reduce the computational footprint of the inverted bottleneck block proposed in ConvNeXt. Our proposed architecture is much less sensitive to evaluation layer count and outperforms a DARTS network with similar size significantly, at layer counts as small as 2. Furthermore, with less layers, not only does it achieve higher accuracy with lower GMACs and parameter count, GradCAM comparisons show that our network is able to better detect distinctive features of target objects compared to DARTS.

Recent advances in language modeling have enabled new conversational systems. In particular, it is often desirable for people to make choices among specified options when using such systems. We address the problem of reference resolution, when people use natural expressions to choose between real world entities. For example, given the choice `Should we make a Simnel cake or a Pandan cake?' a natural response from a non-expert may be indirect: `let's make the green one'. Reference resolution has been little studied with natural expressions, thus robustly understanding such language has large potential for improving naturalness in dialog, recommendation, and search systems. We create AltEntities (Alternative Entities), a new public dataset of entity pairs and utterances, and develop models for the disambiguation problem. Consisting of 42K indirect referring expressions across three domains, it enables for the first time the study of how large language models can be adapted to this task. We find they achieve 82%-87% accuracy in realistic settings, which while reasonable also invites further advances.

The problem of reversing the compilation process, decompilation, is an important tool in reverse engineering of computer software. Recently, researchers have proposed using techniques from neural machine translation to automate the process in decompilation. Although such techniques hold the promise of targeting a wider range of source and assembly languages, to date they have primarily targeted C code. In this paper we argue that existing neural decompilers have achieved higher accuracy at the cost of requiring language-specific domain knowledge such as tokenizers and parsers to build an abstract syntax tree (AST) for the source language, which increases the overhead of supporting new languages. We explore a different tradeoff that, to the extent possible, treats the assembly and source languages as plain text, and show that this allows us to build a decompiler that is easily retargetable to new languages. We evaluate our prototype decompiler, Beyond The C (BTC), on Go, Fortran, OCaml, and C, and examine the impact of parameters such as tokenization and training data selection on the quality of decompilation, finding that it achieves comparable decompilation results to prior work in neural decompilation with significantly less domain knowledge. We will release our training data, trained decompilation models, and code to help encourage future research into language-agnostic decompilation.

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.

Numerous models have tried to effectively embed knowledge graphs in low dimensions. Among the state-of-the-art methods, Graph Neural Network (GNN) models provide structure-aware representations of knowledge graphs. However, they often utilize the information of relations and their interactions with entities inefficiently. Moreover, most state-of-the-art knowledge graph embedding models suffer from scalability issues because of assigning high-dimensional embeddings to entities and relations. To address the above limitations, we propose a scalable general knowledge graph encoder that adaptively involves a powerful tensor decomposition method in the aggregation function of RGCN, a well-known relational GNN model. Specifically, the parameters of a low-rank core projection tensor, used to transform neighborhood entities in the encoder, are shared across relations to benefit from multi-task learning and incorporate relations information effectively. Besides, we propose a low-rank estimation of the core tensor using CP decomposition to compress the model, which is also applicable, as a regularization method, to other similar linear models. We evaluated our model on knowledge graph completion as a common downstream task. We train our model for using a new loss function based on contrastive learning, which relieves the training limitation of the 1-N method on huge graphs. We improved RGCN performance on FB15-237 by 0.42% with considerably lower dimensionality of embeddings.

Gaussian Mixture Models (GMM) are one of the most potent parametric density estimators based on the kernel model that finds application in many scientific domains. In recent years, with the dramatic enlargement of data sources, typical machine learning algorithms, e.g. Expectation Maximization (EM), encounters difficulty with high-dimensional and streaming data. Moreover, complicated densities often demand a large number of Gaussian components. This paper proposes a fast online parameter estimation algorithm for GMM by using first-order stochastic optimization. This approach provides a framework to cope with the challenges of GMM when faced with high-dimensional streaming data and complex densities by leveraging the flexibly-tied factorization of the covariance matrix. A new stochastic Manifold optimization algorithm that preserves the orthogonality is introduced and used along with the well-known Euclidean space numerical optimization. Numerous empirical results on both synthetic and real datasets justify the effectiveness of our proposed stochastic method over EM-based methods in the sense of better-converged maximum for likelihood function, fewer number of needed epochs for convergence, and less time consumption per epoch.

The process of screening molecules for desirable properties is a key step in several applications, ranging from drug discovery to material design. During the process of drug discovery specifically, protein-ligand docking, or chemical docking, is a standard in-silico scoring technique that estimates the binding affinity of molecules with a specific protein target. Recently, however, as the number of virtual molecules available to test has rapidly grown, these classical docking algorithms have created a significant computational bottleneck. We address this problem by introducing Deep Surrogate Docking (DSD), a framework that applies deep learning-based surrogate modeling to accelerate the docking process substantially. DSD can be interpreted as a formalism of several earlier surrogate prefiltering techniques, adding novel metrics and practical training practices. Specifically, we show that graph neural networks (GNNs) can serve as fast and accurate estimators of classical docking algorithms. Additionally, we introduce FiLMv2, a novel GNN architecture which we show outperforms existing state-of-the-art GNN architectures, attaining more accurate and stable performance by allowing the model to filter out irrelevant information from data more efficiently. Through extensive experimentation and analysis, we show that the DSD workflow combined with the FiLMv2 architecture provides a 9.496x speedup in molecule screening with a <3% recall error rate on an example docking task. Our open-source code is available at https://github.com/ryienh/graph-dock.

社会对社交媒体的依赖不断增长，用户为新闻和信息产生的内容增强了不可靠的资源和虚假内容的影响，这使公众讨论并减少了对媒体的信任。验证此类信息的可信度是一项艰巨的任务，容易受到确认偏见的影响，从而开发了算法技术以区分假新闻和真实新闻。但是，大多数现有的方法都具有挑战性的解释，使得难以建立对预测的信任，并在许多现实世界中（例如，视听功能或出处的可用性）做出不现实的假设。在这项工作中，我们专注于使用可解释的功能和方法对文本内容的虚假新闻检测。特别是，我们开发了一个深层的概率模型，该模型使用各种自动编码器和双向长期记忆（LSTM）网络（LSTM）网络与语义主题相关的特征从贝叶斯混合模型推断出来。使用3个现实世界数据集的广泛的实验研究表明，我们的模型可与最先进的竞争模型达到可比的性能，同时促进从学习的主题中解释模型。最后，我们进行了模型消融研究，以证明整合神经嵌入和主题特征的有效性和准确性是通过在较低维嵌入中可分离性评估性能和定性性来定量的。

近年来，视频广播行业一直在显着增长，特别是向最终用户提供个性化内容。虽然视频广播不断增长，但视频纵向已成为直接向观众传递有针对性消息的关键营销工具。但是，不幸的是，对于宽带电视而言，一个关键问题是电视广告针对广泛的受众，因此缺乏特定用户和个性化的广告内容。在本文中，我们提出了一个深云的广告座系统，并简要描述了我们的方法论以及我们设计的广告放置系统的体系结构，以通过MMT流协议交付视频（VOD）和实时广播电视内容。我们论文的目的是展示如何在未来的5G MEC平台上启用针对性，个性化和用户特定的广告服务，这反过来又具有很高的潜力来增加移动运营商行业的广告收入。