Image-text multimodal representation learning aligns data across modalities and enables important medical applications, e.g., image classification, visual grounding, and cross-modal retrieval. In this work, we establish a connection between multimodal representation learning and multiple instance learning. Based on this connection, we propose a generic framework for constructing permutation-invariant score functions with many existing multimodal representation learning approaches as special cases. Furthermore, we use the framework to derive a novel contrastive learning approach and demonstrate that our method achieves state-of-the-art results on a number of downstream tasks.

使用深度学习对胸部射线照相的自动分析具有巨大的潜力，可以增强患者疾病的临床诊断。但是，深度学习模型通常需要大量的带注释的数据来实现高性能 - 通常是医疗领域适应的障碍。在本文中，我们构建了一个利用放射学报告来通过有限的标记数据（少于1000个示例）来改善医学图像分类性能，以提高医学图像分类性能。具体而言，我们检查了捕获图像预告片，以学习以更少的例子进行训练的高质量医学图像表示。在对卷积编码器和变压器解码器进行联合预测之后，我们将学习的编码器转移到各种分类任务中。平均9多种病理学，我们发现我们的模型在标记培训数据受到限制时，比参见和内域监督的预处理的分类性能更高。

在医疗保健中采用机器学习模型需要最终用户对系统的信任。为他们的预测提供额外的支持证据，以促进通过。我们定义了一致的证据，既兼容，又充分涉及模型预测。我们提出了促进更一致的证据的模型不一致和常规方案的措施。我们在从胸部射线照片的水肿严重程度分级的背景下展示了我们的想法。我们经验证明，一致的模型在支持解释时提供竞争性能。

我们提出并通过在图像和文本的本地特征之间最大化互信息来提出并展示表示学习方法。这种方法的目标是通过利用描述图像中发现的自由文本中包含的丰富信息来学习有用的图像表示。我们的方法通过鼓励产生的表示展示了高局部互信息来训练图像和文本编码器。我们利用神经网络鉴别器的互信息估算的最新进展。我们认为，本地互信息的总和通常是全球相互信息的较低限制。我们在下游图像分类任务中的实验结果展示了使用本地特征进行图像文本表示学习的优势。

Reading comprehension of legal text can be a particularly challenging task due to the length and complexity of legal clauses and a shortage of expert-annotated datasets. To address this challenge, we introduce the Merger Agreement Understanding Dataset (MAUD), an expert-annotated reading comprehension dataset based on the American Bar Association's 2021 Public Target Deal Points Study, with over 39,000 examples and over 47,000 total annotations. Our fine-tuned Transformer baselines show promising results, with models performing well above random on most questions. However, on a large subset of questions, there is still room for significant improvement. As the only expert-annotated merger agreement dataset, MAUD is valuable as a benchmark for both the legal profession and the NLP community.

Early detection of relevant locations in a piece of news is especially important in extreme events such as environmental disasters, war conflicts, disease outbreaks, or political turmoils. Additionally, this detection also helps recommender systems to promote relevant news based on user locations. Note that, when the relevant locations are not mentioned explicitly in the text, state-of-the-art methods typically fail to recognize them because these methods rely on syntactic recognition. In contrast, by incorporating a knowledge base and connecting entities with their locations, our system successfully infers the relevant locations even when they are not mentioned explicitly in the text. To evaluate the effectiveness of our approach, and due to the lack of datasets in this area, we also contribute to the research community with a gold-standard multilingual news-location dataset, NewsLOC. It contains the annotation of the relevant locations (and their WikiData IDs) of 600+ Wikinews articles in five different languages: English, French, German, Italian, and Spanish. Through experimental evaluations, we show that our proposed system outperforms the baselines and the fine-tuned version of the model using semi-supervised data that increases the classification rate. The source code and the NewsLOC dataset are publicly available for being used by the research community at https://github.com/vsuarezpaniagua/NewsLocation.

In recent years multi-label, multi-class video action recognition has gained significant popularity. While reasoning over temporally connected atomic actions is mundane for intelligent species, standard artificial neural networks (ANN) still struggle to classify them. In the real world, atomic actions often temporally connect to form more complex composite actions. The challenge lies in recognising composite action of varying durations while other distinct composite or atomic actions occur in the background. Drawing upon the success of relational networks, we propose methods that learn to reason over the semantic concept of objects and actions. We empirically show how ANNs benefit from pretraining, relational inductive biases and unordered set-based latent representations. In this paper we propose deep set conditioned I3D (SCI3D), a two stream relational network that employs latent representation of state and visual representation for reasoning over events and actions. They learn to reason about temporally connected actions in order to identify all of them in the video. The proposed method achieves an improvement of around 1.49% mAP in atomic action recognition and 17.57% mAP in composite action recognition, over a I3D-NL baseline, on the CATER dataset.

Large language models (LLMs) have demonstrated excellent zero-shot generalization to new language tasks. However, effective utilization of LLMs for zero-shot visual question-answering (VQA) remains challenging, primarily due to the modality disconnection and task disconnection between LLM and VQA task. End-to-end training on vision and language data may bridge the disconnections, but is inflexible and computationally expensive. To address this issue, we propose \emph{Img2Prompt}, a plug-and-play module that provides the prompts that can bridge the aforementioned modality and task disconnections, so that LLMs can perform zero-shot VQA tasks without end-to-end training. In order to provide such prompts, we further employ LLM-agnostic models to provide prompts that can describe image content and self-constructed question-answer pairs, which can effectively guide LLM to perform zero-shot VQA tasks. Img2Prompt offers the following benefits: 1) It can flexibly work with various LLMs to perform VQA. 2)~Without the needing of end-to-end training, it significantly reduces the cost of deploying LLM for zero-shot VQA tasks. 3) It achieves comparable or better performance than methods relying on end-to-end training. For example, we outperform Flamingo~\cite{Deepmind:Flamingo2022} by 5.6\% on VQAv2. On the challenging A-OKVQA dataset, our method even outperforms few-shot methods by as much as 20\%.

As information extraction (IE) systems have grown more capable at whole-document extraction, the classic task of \emph{template filling} has seen renewed interest as a benchmark for evaluating them. In this position paper, we call into question the suitability of template filling for this purpose. We argue that the task demands definitive answers to thorny questions of \emph{event individuation} -- the problem of distinguishing distinct events -- about which even human experts disagree. We show through annotation studies and error analysis that this raises concerns about the usefulness of template filling evaluation metrics, the quality of datasets for the task, and the ability of models to learn it. Finally, we consider possible solutions.

Machine learning models are increasingly deployed for critical decision-making tasks, making it important to verify that they do not contain gender or racial biases picked up from training data. Typical approaches to achieve fairness revolve around efforts to clean or curate training data, with post-hoc statistical evaluation of the fairness of the model on evaluation data. In contrast, we propose techniques to \emph{prove} fairness using recently developed formal methods that verify properties of neural network models.Beyond the strength of guarantee implied by a formal proof, our methods have the advantage that we do not need explicit training or evaluation data (which is often proprietary) in order to analyze a given trained model. In experiments on two familiar datasets in the fairness literature (COMPAS and ADULTS), we show that through proper training, we can reduce unfairness by an average of 65.4\% at a cost of less than 1\% in AUC score.