In this paper, we apply transformer-based Natural Language Generation (NLG) techniques to the problem of text simplification. Currently, there are only a few German datasets available for text simplification, even fewer with larger and aligned documents, and not a single one with narrative texts. In this paper, we explore to which degree modern NLG techniques can be applied to German narrative text simplifications. We use Longformer attention and a pre-trained mBART model. Our findings indicate that the existing approaches for German are not able to solve the task properly. We conclude on a few directions for future research to address this problem.
We are witnessing significant breakthroughs in the technology for generating 3D objects from text. Existing approaches either leverage large text-to-image models to optimize a 3D representation or train 3D generators on object-centric datasets. Generating entire scenes, however, remains very challenging as a scene contains multiple 3D objects, diverse and scattered. In this work, we introduce SceneWiz3D, a novel approach to synthesize high-fidelity 3D scenes from text. We marry the locality of objects with globality of scenes by introducing a hybrid 3D representation: explicit for objects and implicit for scenes. Remarkably, an object, being represented explicitly, can be either generated from text using conventional text-to-3D approaches, or provided by users. To configure the layout of the scene and automatically place objects, we apply the Particle Swarm Optimization technique during the optimization process. Furthermore, it is difficult for certain parts of the scene (e.g., corners, occlusion) to receive multi-view supervision, leading to inferior geometry. We incorporate an RGBD panorama diffusion model to mitigate it, resulting in high-quality geometry. Extensive evaluation supports that our approach achieves superior quality over previous approaches, enabling the generation of detailed and view-consistent 3D scenes.
Objective To solve major clinical natural language processing (NLP) tasks using a unified text-to-text learning architecture based on a generative large language model (LLM) via prompt tuning. Methods We formulated 7 key clinical NLP tasks as text-to-text learning and solved them using one unified generative clinical LLM, GatorTronGPT, developed using GPT-3 architecture and trained with up to 20 billion parameters. We adopted soft prompts (i.e., trainable vectors) with frozen LLM, where the LLM parameters were not updated (i.e., frozen) and only the vectors of soft prompts were updated, known as prompt tuning. We added additional soft prompts as a prefix to the input layer, which were optimized during the prompt tuning. We evaluated the proposed method using 7 clinical NLP tasks and compared them with previous task-specific solutions based on Transformer models. Results and Conclusion The proposed approach achieved state-of-the-art performance for 5 out of 7 major clinical NLP tasks using one unified generative LLM. Our approach outperformed previous task-specific transformer models by ~3% for concept extraction and 7% for relation extraction applied to social determinants of health, 3.4% for clinical concept normalization, 3.4~10% for clinical abbreviation disambiguation, and 5.5~9% for natural language inference. Our approach also outperformed a previously developed prompt-based machine reading comprehension (MRC) model, GatorTron-MRC, for clinical concept and relation extraction. The proposed approach can deliver the ``one model for all`` promise from training to deployment using a unified generative LLM.
As legal case law databases such as HUDOC continue to grow rapidly, it has become essential for legal researchers to find efficient methods to handle such large-scale data sets. Such case law databases usually consist of the textual content of cases together with the citations between them. This paper focuses on case law from the European Court of Human Rights on Article 8 of the European Convention of Human Rights, the right to respect private and family life, home and correspondence. In this study, we demonstrate and compare the potential of topic modelling and citation network to find and organize case law on Article 8 based on their general themes and citation patterns, respectively. Additionally, we explore whether combining these two techniques leads to better results compared to the application of only one of the methods. We evaluate the effectiveness of the combined method on a unique manually collected and annotated dataset of Aricle 8 case law on evictions. The results of our experiments show that our combined (text and citation-based) approach provides the best results in finding and grouping case law, providing scholars with an effective way to extract and analyse relevant cases on a specific issue.
Visible-infrared person re-identification (VIReID) primarily deals with matching identities across person images from different modalities. Due to the modality gap between visible and infrared images, cross-modality identity matching poses significant challenges. Recognizing that high-level semantics of pedestrian appearance, such as gender, shape, and clothing style, remain consistent across modalities, this paper intends to bridge the modality gap by infusing visual features with high-level semantics. Given the capability of CLIP to sense high-level semantic information corresponding to visual representations, we explore the application of CLIP within the domain of VIReID. Consequently, we propose a CLIP-Driven Semantic Discovery Network (CSDN) that consists of Modality-specific Prompt Learner, Semantic Information Integration (SII), and High-level Semantic Embedding (HSE). Specifically, considering the diversity stemming from modality discrepancies in language descriptions, we devise bimodal learnable text tokens to capture modality-private semantic information for visible and infrared images, respectively. Additionally, acknowledging the complementary nature of semantic details across different modalities, we integrate text features from the bimodal language descriptions to achieve comprehensive semantics. Finally, we establish a connection between the integrated text features and the visual features across modalities. This process embed rich high-level semantic information into visual representations, thereby promoting the modality invariance of visual representations. The effectiveness and superiority of our proposed CSDN over existing methods have been substantiated through experimental evaluations on multiple widely used benchmarks. The code will be released at \url{https://github.com/nengdong96/CSDN}.
Federated Learning (FL) addresses critical issues in machine learning related to data privacy and security, yet suffering from data insufficiency and imbalance under certain circumstances. The emergence of foundation models (FMs) offers potential solutions to the limitations of existing FL frameworks, e.g., by generating synthetic data for model initialization. However, due to the inherent safety concerns of FMs, integrating FMs into FL could introduce new risks, which remains largely unexplored. To address this gap, we conduct the first investigation on the vulnerability of FM integrated FL (FM-FL) under adversarial threats. Based on a unified framework of FM-FL, we introduce a novel attack strategy that exploits safety issues of FM to compromise FL client models. Through extensive experiments with well-known models and benchmark datasets in both image and text domains, we reveal the high susceptibility of the FM-FL to this new threat under various FL configurations. Furthermore, we find that existing FL defense strategies offer limited protection against this novel attack approach. This research highlights the critical need for enhanced security measures in FL in the era of FMs.
In recent years, image editing has advanced remarkably. With increased human control, it is now possible to edit an image in a plethora of ways; from specifying in text what we want to change, to straight up dragging the contents of the image in an interactive point-based manner. However, most of the focus has remained on editing single images at a time. Whether and how we can simultaneously edit large batches of images has remained understudied. With the goal of minimizing human supervision in the editing process, this paper presents a novel method for interactive batch image editing using StyleGAN as the medium. Given an edit specified by users in an example image (e.g., make the face frontal), our method can automatically transfer that edit to other test images, so that regardless of their initial state (pose), they all arrive at the same final state (e.g., all facing front). Extensive experiments demonstrate that edits performed using our method have similar visual quality to existing single-image-editing methods, while having more visual consistency and saving significant time and human effort.
Recent models have achieved human level performance on the Stanford Question Answering Dataset when using F1 scores to evaluate the reading comprehension task. Yet, teaching machines to comprehend text has not been solved in the general case. By appending one adversarial sentence to the context paragraph, past research has shown that the F1 scores from reading comprehension models drop almost in half. In this paper, I replicate past adversarial research with a new model, ELECTRA-Small, and demonstrate that the new model's F1 score drops from 83.9% to 29.2%. To improve ELECTRA-Small's resistance to this attack, I finetune the model on SQuAD v1.1 training examples with one to five adversarial sentences appended to the context paragraph. Like past research, I find that the finetuned model on one adversarial sentence does not generalize well across evaluation datasets. However, when finetuned on four or five adversarial sentences the model attains an F1 score of more than 70% on most evaluation datasets with multiple appended and prepended adversarial sentences. The results suggest that with enough examples we can make models robust to adversarial attacks.
World models play a crucial role in understanding and predicting the dynamics of the world, which is essential for video generation. However, existing world models are confined to specific scenarios such as gaming or driving, limiting their ability to capture the complexity of general world dynamic environments. Therefore, we introduce WorldDreamer, a pioneering world model to foster a comprehensive comprehension of general world physics and motions, which significantly enhances the capabilities of video generation. Drawing inspiration from the success of large language models, WorldDreamer frames world modeling as an unsupervised visual sequence modeling challenge. This is achieved by mapping visual inputs to discrete tokens and predicting the masked ones. During this process, we incorporate multi-modal prompts to facilitate interaction within the world model. Our experiments show that WorldDreamer excels in generating videos across different scenarios, including natural scenes and driving environments. WorldDreamer showcases versatility in executing tasks such as text-to-video conversion, image-tovideo synthesis, and video editing. These results underscore WorldDreamer's effectiveness in capturing dynamic elements within diverse general world environments.
Background: The COVID-19 pandemic has caused severe impacts on health systems worldwide. Its critical nature and the increased interest of individuals and organizations to develop countermeasures to the problem has led to a surge of new studies in scientific journals. Objetive: We sought to develop a tool that incorporates, in a novel way, aspects of Information Retrieval (IR) and Extraction (IE) applied to the COVID-19 Open Research Dataset (CORD-19). The main focus of this paper is to provide researchers with a better search tool for COVID-19 related papers, helping them find reference papers and hightlight relevant entities in text. Method: We applied Latent Dirichlet Allocation (LDA) to model, based on research aspects, the topics of all English abstracts in CORD-19. Relevant named entities of each abstract were extracted and linked to the corresponding UMLS concept. Regular expressions and the K-Nearest Neighbors algorithm were used to rank relevant papers. Results: Our tool has shown the potential to assist researchers by automating a topic-based search of CORD-19 papers. Nonetheless, we identified that more fine-tuned topic modeling parameters and increased accuracy of the research aspect classifier model could lead to a more accurate and reliable tool. Conclusion: We emphasize the need of new automated tools to help researchers find relevant COVID-19 documents, in addition to automatically extracting useful information contained in them. Our work suggests that combining different algorithms and models could lead to new ways of browsing COVID-19 paper data.