Machine-generated text is increasingly difficult to distinguish from text authored by humans. Powerful open-source models are freely available, and user-friendly tools that democratize access to generative models are proliferating. ChatGPT, which was released shortly after the first edition of this survey, epitomizes these trends. The great potential of state-of-the-art natural language generation (NLG) systems is tempered by the multitude of avenues for abuse. Detection of machine-generated text is a key countermeasure for reducing the abuse of NLG models, and presents significant technical challenges and numerous open problems. We provide a survey that includes 1) an extensive analysis of threat models posed by contemporary NLG systems and 2) the most complete review of machine-generated text detection methods to date. This survey places machine-generated text within its cybersecurity and social context, and provides strong guidance for future work addressing the most critical threat models. While doing so, we highlight the importance that detection systems themselves demonstrate trustworthiness through fairness, robustness, and accountability.

Machine-Generated Text: A Comprehensive Survey of Threat Models and Detection Methods

Text-based Visual Question Answering~(TextVQA) aims to produce correct answers for given questions about the images with multiple scene texts. In most cases, the texts naturally attach to the surface of the objects. Therefore, spatial reasoning between texts and objects is crucial in TextVQA. However, existing approaches are constrained within 2D spatial information learned from the input images and rely on transformer-based architectures to reason implicitly during the fusion process. Under this setting, these 2D spatial reasoning approaches cannot distinguish the fine-grain spatial relations between visual objects and scene texts on the same image plane, thereby impairing the interpretability and performance of TextVQA models. In this paper, we introduce 3D geometric information into a human-like spatial reasoning process to capture the contextual knowledge of key objects step-by-step. %we formulate a human-like spatial reasoning process by introducing 3D geometric information for capturing key objects' contextual knowledge. To enhance the model's understanding of 3D spatial relationships, Specifically, (i)~we propose a relation prediction module for accurately locating the region of interest of critical objects; (ii)~we design a depth-aware attention calibration module for calibrating the OCR tokens' attention according to critical objects. Extensive experiments show that our method achieves state-of-the-art performance on TextVQA and ST-VQA datasets. More encouragingly, our model surpasses others by clear margins of 5.7\% and 12.1\% on questions that involve spatial reasoning in TextVQA and ST-VQA valid split. Besides, we also verify the generalizability of our model on the text-based image captioning task.

Toward 3D Spatial Reasoning for Human-like Text-based Visual Question Answering

Compositional zero-shot learning aims to recognize unseen compositions of seen visual primitives of object classes and their states. While all primitives (states and objects) are observable during training in some combination, their complex interaction makes this task especially hard. For example, wet changes the visual appearance of a dog very differently from a bicycle. Furthermore, we argue that relationships between compositions go beyond shared states or objects. A cluttered office can contain a busy table; even though these compositions don’t share a state or object, the presence of a busy table can guide the presence of a cluttered office. We propose a novel method called Compositional Attention Propagated Embedding (CAPE) as a solution. The key intuition to our method is that a rich dependency structure exists between compositions arising from complex interactions of primitives in addition to other dependencies between compositions. CAPE learns to identify this structure and propagates knowledge between them to learn class embedding for all seen and unseen compositions. In the challenging generalized compositional zero-shot setting, we show that our method outperforms previous baselines to set a new state-of-the-art on three publicly available benchmarks.

Learning Attention Propagation for Compositional Zero-Shot Learning

The goal of multimodal summarization is to extract the most important information from different modalities to form output summaries. Unlike the unimodal summarization, the multimodal summarization task explicitly leverages cross-modal information to help generate more reliable and high-quality summaries. However, existing methods fail to leverage the temporal correspondence between different modalities and ignore the intrinsic correlation between different samples. To address this issue, we introduce Align and Attend Multimodal Summarization (A2Summ), a unified multimodal transformer-based model which can effectively align and attend the multimodal input. In addition, we propose two novel contrastive losses to model both inter-sample and intra-sample correlations. Extensive experiments on two standard video summarization datasets (TVSum and SumMe) and two multimodal summarization datasets (Daily Mail and CNN) demonstrate the superiority of A2Summ, achieving state-of-the-art performances on all datasets. Moreover, we collected a large-scale multimodal summarization dataset BLiSS, which contains livestream videos and transcribed texts with annotated summaries. Our code and dataset are publicly available at ~\url{https://boheumd.github.io/A2Summ/}.

Align and Attend: Multimodal Summarization with Dual Contrastive Losses

We propose DocFormerv2, a multi-modal transformer for Visual Document Understanding (VDU). The VDU domain entails understanding documents (beyond mere OCR predictions) e.g., extracting information from a form, VQA for documents and other tasks. VDU is challenging as it needs a model to make sense of multiple modalities (visual, language and spatial) to make a prediction. Our approach, termed DocFormerv2 is an encoder-decoder transformer which takes as input - vision, language and spatial features. DocFormerv2 is pre-trained with unsupervised tasks employed asymmetrically i.e., two novel document tasks on encoder and one on the auto-regressive decoder. The unsupervised tasks have been carefully designed to ensure that the pre-training encourages local-feature alignment between multiple modalities. DocFormerv2 when evaluated on nine datasets shows state-of-the-art performance over strong baselines e.g. TabFact (4.3%), InfoVQA (1.4%), FUNSD (1%). Furthermore, to show generalization capabilities, on three VQA tasks involving scene-text, Doc- Formerv2 outperforms previous comparably-sized models and even does better than much larger models (such as GIT2, PaLi and Flamingo) on some tasks. Extensive ablations show that due to its pre-training, DocFormerv2 understands multiple modalities better than prior-art in VDU.

DocFormerv2: Local Features for Document Understanding

Text-VQA aims at answering questions that require understanding the textual cues in an image. Despite the great progress of existing Text-VQA methods, their performance suffers from insufficient human-labeled question-answer (QA) pairs. However, we observe that, in general, the scene text is not fully exploited in the existing datasets -- only a small portion of the text in each image participates in the annotated QA activities. This results in a huge waste of useful information. To address this deficiency, we develop a new method to generate high-quality and diverse QA pairs by explicitly utilizing the existing rich text available in the scene context of each image. Specifically, we propose, TAG, a text-aware visual question-answer generation architecture that learns to produce meaningful, and accurate QA samples using a multimodal transformer. The architecture exploits underexplored scene text information and enhances scene understanding of Text-VQA models by combining the generated QA pairs with the initial training data. Extensive experimental results on two well-known Text-VQA benchmarks (TextVQA and ST-VQA) demonstrate that our proposed TAG effectively enlarges the training data that helps improve the Text-VQA performance without extra labeling effort. Moreover, our model outperforms state-of-the-art approaches that are pre-trained with extra large-scale data. Code is available at https://github.com/HenryJunW/TAG.

TAG: Boosting Text-VQA via Text-aware Visual Question-answer Generation

Objectives The foveal avascular zone (FAZ) is a biomarker for quantifying diabetic macular ischemia (DMI), to automate the identification and quantification of the FAZ in DMI, using an improved U-Net convolutional neural network (CNN) and to establish a CNN model based on optical coherence tomography angiography (OCTA) images for the same purpose. Methods The FAZ boundaries on the full-thickness retina of 6 × 6 mm en face OCTA images of DMI and normal eyes were manually marked. Seventy percent of OCTA images were used as the training set, and ten percent of these images were used as the validation set to train the improved U-Net CNN with two attention modules. Finally, twenty percent of the OCTA images were used as the test set to evaluate the accuracy of this model relative to that of the baseline U-Net model. This model was then applied to the public data set sFAZ to compare its effectiveness with existing models at identifying and quantifying the FAZ area. Results This study included 110 OCTA images. The Dice score of the FAZ area predicted by the proposed method was 0.949, the Jaccard index was 0.912, and the area correlation coefficient was 0.996. The corresponding values for the baseline U-Net were 0.940, 0.898, and 0.995, respectively, and those based on the description data set sFAZ were 0.983, 0.968, and 0.950, respectively, which were better than those previously reported based on this data set. Conclusions The improved U-Net CNN was more accurate at automatically measuring the FAZ area on the OCTA images than the traditional CNN. The present model may measure the DMI index more accurately, thereby assisting in the diagnosis and prognosis of retinal vascular diseases such as diabetic retinopathy.

Yuqian Hu

Papers

TAG: Boosting Text-VQA via Text-aware Visual Question-answer Generation

Application of Improved U-Net Convolutional Neural Network for Automatic Quantification of the Foveal Avascular Zone in Diabetic Macular Ischemia