论文检索

Yanbei Jiang, Krista A. Ehinger, Jey Han Lau

Exploring the narratives conveyed by fine-art paintings is a challenge in image captioning, where the goal is to generate descriptions that not only precisely represent the visual content but also offer a in-depth interpretation of the artwork's meaning. The task is particularly complex for artwork images due to their diverse interpretations and varied aesthetic principles across different artistic schools and styles. In response to this, we present KALE (Knowledge-Augmented vision-Language model for artwork Elaborations), a novel approach that enhances existing vision-language models by integrating artwork metadata as additional knowledge. KALE incorporates the metadata in two ways: firstly as direct textual input, and secondly through a multimodal heterogeneous knowledge graph. To optimize the learning of graph representations, we introduce a new cross-modal alignment loss that maximizes the similarity between the image and its corresponding metadata. Experimental results demonstrate that KALE achieves strong performance (when evaluated with CIDEr, in particular) over existing state-of-the-art work across several artwork datasets. Source code of the project is available at https://github.com/Yanbei-Jiang/Artwork-Interpretation.

Chongyang Gao, Yiren Jian, Natalia Denisenko, Soroush Vosoughi, V. S. Subrahmanian

Generating engaging multimodal content is a key objective in numerous applications, such as the creation of online advertisements that captivate user attention through a synergy of images and text. In this paper, we introduce GEM, a novel framework engineered for the generation of engaging multimodal image-text posts. The GEM framework operates in two primary phases. Initially, GEM integrates a pre-trained engagement discriminator with a technique for deriving an effective continuous prompt tailored for the stable diffusion model. Subsequently, GEM unveils an iterative algorithm dedicated to producing coherent and compelling image-sentence pairs centered around a specified topic of interest. Through a combination of experimental analysis and human evaluations, we establish that the image-sentence pairs generated by GEM not only surpass several established baselines in terms of engagement but also in achieving superior alignment.

Zhongjie Duan, Chengyu Wang, Cen Chen, Weining Qian, Jun Huang

Toon shading is a type of non-photorealistic rendering task in animation. Its primary purpose is to render objects with a flat and stylized appearance. As diffusion models have ascended to the forefront of image synthesis, this paper delves into an innovative form of toon shading based on diffusion models, aiming to directly render photorealistic videos into anime styles. In video stylization, existing methods encounter persistent challenges, notably in maintaining consistency and achieving high visual quality. In this paper, we model the toon shading problem as four subproblems, i.e., stylization, consistency enhancement, structure guidance, and colorization. To address the challenges in video stylization, we propose an effective toon shading approach called Diffutoon. Diffutoon is capable of rendering remarkably detailed, high-resolution, and extended-duration videos in anime style. It can also edit the video content according to input prompts via an additional branch. The efficacy of Diffutoon is evaluated through quantitive metrics and human evaluation. Notably, Diffutoon surpasses both open-source and closed-source baseline approaches in our experiments. Our work is accompanied by the release of both the source code and example videos on Github.

Niklas Deckers, Julia Peters, Martin Potthast

Prompt engineering is still the primary way for users of generative text-to-image models to manipulate generated images in a targeted way. Based on treating the model as a continuous function and by passing gradients between the image space and the prompt embedding space, we propose and analyze a new method to directly manipulate the embedding of a prompt instead of the prompt text. We then derive three practical interaction tools to support users with image generation: (1) Optimization of a metric defined in the image space that measures, for example, the image style. (2) Supporting a user in creative tasks by allowing them to navigate in the image space along a selection of directions of "near" prompt embeddings. (3) Changing the embedding of the prompt to include information that a user has seen in a particular seed but has difficulty describing in the prompt. Compared to prompt engineering, user-driven prompt embedding manipulation enables a more fine-grained, targeted control that integrates a user's intentions. Our user study shows that our methods are considered less tedious and that the resulting images are often preferred.

Allegra De Filippo, Michela Milano

This paper focuses on the potential of Generative Artificial Intelligence (AI), particularly Large Language Models (LLMs), in the still unexplored domain of robotic dance creation. In particular, we assess whether a LLM (GPT-3.5 turbo) can create robotic dance choreographies, and we investigate if the feedback provided by human creators can improve the quality of the output. To this end, we design three prompt engineering techniques for robotic dance creation. In the prompts, we gradually introduce human knowledge through examples and feedback in natural language in order to explore the dynamics of human-AI co-creation. The experimental analysis shows that the capabilities of the LLM can be improved through human collaboration, by producing choreographies with a major artistic impact on the evaluation audience. The findings offer valuable insights into the interplay between human creativity and AI generative models, paving the way for enhanced collaborative frameworks in creative domains.

Jianyi Chen, Wei Xue, Xu Tan, Zhen Ye, Qifeng Liu, Yike Guo

Singing Accompaniment Generation (SAG), which generates instrumental music to accompany input vocals, is crucial to developing human-AI symbiotic art creation systems. The state-of-the-art method, SingSong, utilizes a multi-stage autoregressive (AR) model for SAG, however, this method is extremely slow as it generates semantic and acoustic tokens recursively, and this makes it impossible for real-time applications. In this paper, we aim to develop a Fast SAG method that can create high-quality and coherent accompaniments. A non-AR diffusion-based framework is developed, which by carefully designing the conditions inferred from the vocal signals, generates the Mel spectrogram of the target accompaniment directly. With diffusion and Mel spectrogram modeling, the proposed method significantly simplifies the AR token-based SingSong framework, and largely accelerates the generation. We also design semantic projection, prior projection blocks as well as a set of loss functions, to ensure the generated accompaniment has semantic and rhythm coherence with the vocal signal. By intensive experimental studies, we demonstrate that the proposed method can generate better samples than SingSong, and accelerate the generation by at least 30 times. Audio samples and code are available at this link.

João Libório Cardoso, Francesco Banterle, Paolo Cignoni, Michael Wimmer

We introduce context-aware translation, a novel method that combines the benefits of inpainting and image-to-image translation, respecting simultaneously the original input and contextual relevance -- where existing methods fall short. By doing so, our method opens new avenues for the controllable use of AI within artistic creation, from animation to digital art. As an use case, we apply our method to redraw any hand-drawn animated character eyes based on any design specifications -- eyes serve as a focal point that captures viewer attention and conveys a range of emotions; however, the labor-intensive nature of traditional animation often leads to compromises in the complexity and consistency of eye design. Furthermore, we remove the need for production data for training and introduce a new character recognition method that surpasses existing work by not requiring fine-tuning to specific productions. This proposed use case could help maintain consistency throughout production and unlock bolder and more detailed design choices without the production cost drawbacks. A user study shows context-aware translation is preferred over existing work 95.16% of the time.

Alexandre Bonlarron, Jean-Charles Régin

Constrained text generation remains a challenging task, particularly when dealing with hard constraints. Traditional Natural Language Processing (NLP) approaches prioritize generating meaningful and coherent output. Also, the current state-of-the-art methods often lack the expressiveness and constraint satisfaction capabilities to handle such tasks effectively. This paper presents the Constraints First Framework to remedy this issue. This framework considers a constrained text generation problem as a discrete combinatorial optimization problem. It is solved by a constraint programming method that combines linguistic properties (e.g., n-grams or language level) with other more classical constraints (e.g., the number of characters, syllables, or words). Eventually, a curation phase allows for selecting the best-generated sentences according to perplexity using a large language model. The effectiveness of this approach is demonstrated by tackling a new more tediously constrained text generation problem: the iconic RADNER sentences problem. This problem aims to generate sentences respecting a set of quite strict rules defined by their use in vision and clinical research. Thanks to our CP-based approach, many new strongly constrained sentences have been successfully generated in an automatic manner. This highlights the potential of our approach to handle unreasonably constrained text generation scenarios.

Jinbin Bai, Zhen Dong, Aosong Feng, Xiao Zhang, Tian Ye, Kaicheng Zhou

In the field of image processing, applying intricate semantic modifications within existing images remains an enduring challenge. This paper introduces a pioneering framework that integrates viewpoint information to enhance the control of image editing tasks, especially for interior design scenes. By surveying existing object editing methodologies, we distill three essential criteria --- consistency, controllability, and harmony --- that should be met for an image editing method. In contrast to previous approaches, our framework takes the lead in satisfying all three requirements for addressing the challenge of image synthesis. Through comprehensive experiments, encompassing both quantitative assessments and qualitative comparisons with contemporary state-of-the-art methods, we present compelling evidence of our framework's superior performance across multiple dimensions. This work establishes a promising avenue for advancing image synthesis techniques and empowering precise object modifications while preserving the visual coherence of the entire composition.

Yao Zou, Dawei Cheng

Credit card fraud imposes significant costs on both cardholders and issuing banks. Fraudsters often disguise their crimes, such as using legitimate transactions through several benign users to bypass anti-fraud detection. Existing graph neural network (GNN) models struggle with learning features of camouflaged, indirect multi-hop transactions due to their inherent over-smoothing issues in deep multi-layer aggregation, presenting a major challenge in detecting disguised relationships. Therefore, in this paper, we propose a novel High-order Graph Representation Learning model (HOGRL) to avoid incorporating excessive noise during the multi-layer aggregation process. In particular, HOGRL learns different orders of \emph{pure} representations directly from high-order transaction graphs. We realize this goal by effectively constructing high-order transaction graphs first and then learning the \emph{pure} representations of each order so that the model could identify fraudsters' multi-hop indirect transactions via multi-layer \emph{pure} feature learning. In addition, we introduce a mixture-of-expert attention mechanism to automatically determine the importance of different orders for jointly optimizing fraud detection performance. We conduct extensive experiments in both the open source and real-world datasets, the result demonstrates the significant improvements of our proposed HOGRL compared with state-of-the-art fraud detection baselines. HOGRL's superior performance also proves its effectiveness in addressing high-order fraud camouflage criminals.

Menghao Zhang, Jingyu Wang, Qi Qi, Pengfei Ren, Haifeng Sun, Zirui Zhuang, Lei Zhang, Jianxin Liao

Sustainable cities requires high-quality community management and surveillance analytics, which are supported by video anomaly detection techniques. However, mainstream video anomaly detection techniques still require manually labeled data and do not apply to real-world massive videos. Without labeling, unsupervised video anomaly detection (UVAD) is challenged by the problem of pseudo-labeled noise and the openness of anomaly detection. In response, a diffusion-based latent pattern learning UVAD framework is proposed, called DiffVAD. The method learns potential patterns by generating different patterns of the same event through diffusion models. The detection of anomalies is realized by evaluating the pattern distribution. The different patterns of normal events are diverse but correlated, while the different patterns of abnormal events are more diffuse. This manner of detection is equally effective for unseen normal events in the training set. In addition, we design a refinement strategy for pseudo-labels to mitigate the effects of the noise problem. Extensive experiments on six benchmark datasets demonstrate the design’s promising generalization ability and high efficiency. Specifically, DiffVAD obtains an AUC score of 81.9% on the ShanghaiTech dataset.

Lixian Zhang, Runmin Dong, Shuai Yuan, Jinxiao Zhang, Mengxuan Chen, Juepeng Zheng, Haohuan Fu

Nighttime light (NTL) remote sensing observation serves as a unique proxy for quantitatively assessing progress toward meeting a series of Sustainable Development Goals (SDGs), such as poverty estimation, urban sustainable development, and carbon emission. However, existing NTL observations often suffer from pervasive degradation and inconsistency, limiting their utility for computing the indicators defined by the SDGs. In this study, we propose a novel approach to reconstruct high-resolution NTL images using multimodal remote sensing data. To support this research endeavor, we introduce DeepLightMD, a comprehensive dataset comprising data from five heterogeneous sensors, offering fine spatial resolution and rich spectral information at a national scale. Additionally, we present DeepLightSR, a calibration-aware method for multi-modality super-resolution. DeepLightSR integrates calibration-aware alignment, an auxiliary-to-main multi-modality fusion, and an auxiliary-embedded refinement to effectively address spatial heterogeneity, fuse diversely representative features, and enhance performance in ×8 super-resolution tasks. Extensive experiments demonstrate the superiority of DeepLightSR over 8 competing methods, as evidenced by improvements in PSNR (2.01 dB ~ 13.25 dB) and PIQE (0.49 ~ 9.32). Our findings underscore the practical significance of our proposed dataset and model in reconstructing high-resolution NTL data, supporting efficiently and quantitatively assessing the SDG progress. The code and data will be available at https://github.com/xian1234/DeepLight.

Huaiwu Zhang, Yutong Xia, Siru Zhong, Kun Wang, Zekun Tong, Qingsong Wen, Roger Zimmermann, Yuxuan Liang

The increasing number of vehicles highlights the need for efficient parking space management. Predicting real-time Parking Availability (PA) can help mitigate traffic congestion and the corresponding social problems, which is a pressing issue in densely populated cities like Singapore. In this study, we aim to collectively predict future PA across Singapore with complex factors from various domains. The contributions in this paper are listed as follows: (1) A New Dataset: We introduce the SINPA dataset, containing a year's worth of PA data from 1,687 parking lots in Singapore, enriched with various spatial and temporal factors. (2) A Data-Driven Approach: We present DeepPA, a novel deep-learning framework, to collectively and efficiently predict future PA across thousands of parking lots. (3) Extensive Experiments and Deployment: DeepPA demonstrates a 9.2% reduction in prediction error for up to 3-hour forecasts compared to existing advanced models. Furthermore, we implement DeepPA in a practical web-based platform to provide real-time PA predictions to aid drivers and inform urban planning for the governors in Singapore. We release the dataset and source code at https://github.com/yoshall/SINPA.

Zijie Zeng, Shiqi Liu, Lele Sha, Zhuang Li, Kaixun Yang, Sannyuya Liu, Dragan Gasevic, Guangliang Chen

This study explores the challenge of sentence-level AI-generated text detection within human-AI collaborative hybrid texts (abbreviated as hybrid texts). Existing studies of AI-generated text detection for hybrid texts often rely on synthetic datasets. These typically involve hybrid texts with a limited number of boundaries, e.g., single-boundary hybrid texts that begin with human-written content and end with machine-generated continuations. We contend that studies of detecting AI-generated content within hybrid texts should cover different types of hybrid texts generated in realistic settings to better inform real-world applications. Therefore, our study utilizes the CoAuthor dataset, which includes diverse, realistic hybrid texts generated through the collaboration between human writers and an intelligent writing system in multi-turn interactions. We adopt a two-step, segmentation-based pipeline: (i) detect segments within a given hybrid text where each segment contains sentences of consistent authorship, and (ii) classify the authorship of each identified segment. Our empirical findings highlight (1) detecting AI-generated sentences in hybrid texts is overall a challenging task because (1.1) human writers' selecting and even editing AI-generated sentences based on personal preferences adds difficulty in identifying the authorship of segments; (1.2) the frequent change of authorship between neighboring sentences within the hybrid text creates difficulties for segment detectors in identifying authorship-consistent segments; (1.3) the short length of text segments within hybrid texts provides limited stylistic cues for reliable authorship determination; (2) before embarking on the detection process, it is beneficial to assess the average length of segments within the hybrid text. This assessment aids in deciding whether (2.1) to employ a text segmentation-based strategy for hybrid texts with longer segments, or (2.2) to adopt a direct sentence-by-sentence classification strategy for those with shorter segments.

Xixian Yong, Xiao Zhou

Predicting socioeconomic indicators within urban regions is crucial for fostering inclusivity, resilience, and sustainability in cities and human settlements. While pioneering studies have attempted to leverage multi-modal data for socioeconomic prediction, jointly exploring their underlying semantics remains a significant challenge. To address the gap, this paper introduces a Multi-Semantic Contrastive Learning (MuseCL) framework for fine-grained urban region profiling and socioeconomic prediction. Within this framework, we initiate the process by constructing contrastive sample pairs for street view and remote sensing images, capitalizing on the similarities in human mobility and Point of Interest (POI) distribution to derive semantic features from the visual modality. Additionally, we extract semantic insights from POI texts embedded within these regions, employing a pre-trained text encoder. To merge the acquired visual and textual features, we devise an innovative cross-modality-based attentional fusion module, which leverages a contrastive mechanism for integration. Experimental results across multiple cities and indicators consistently highlight the superiority of MuseCL, demonstrating an average improvement of 10% in R2 compared to various competitive baseline models. The code of this work is publicly available at https://github.com/XixianYong/MuseCL.

Zongyue Xue, Huanghai Liu, Yiran Hu, Yuliang Qian, Yajing Wang, Kangle Kong, Chenlu Wang, Yun Liu, Weixing Shen

An extensive label system is pivotal to facilitate judicial fairness and social justice. Prior empirical research and our interview with legal professionals underscore the importance of extra-legal factors in criminal trials. To help identify sentencing biases and facilitate downstream applications, we introduce the Legal Element ExtraCtion (LEEC) dataset comprising 15,919 judicial documents and 155 labels. This dataset was constructed through two main steps: First, designing the label system by legal experts based on prior empirical research which identified critical factors driving and processes generating sentencing outcomes in criminal cases; Second, employing legal knowledge to annotate judicial documents according to the label system and annotation guideline. LEEC represents the most extensive and domain-specific legal element extraction dataset for the Chinese legal system. Our experiments reveal that despite certain capabilities, both Document Event Extraction (DEE) models and Large Language Models(LLMs) face significant restrictions in legal element extraction tasks. Finally, our empirical analysis based on LEEC provides evidence for judicial unfairness in Chinese criminal sentencing and confirms the applicability of LEEC for future empirical research and other downstream applications. LEEC and related resources are available on https://github.com/THUlawtech/LEEC.

Zhuo Xu, Xiao Zhou

The explosion of massive urban data recently has provided us with a valuable opportunity to gain deeper insights into urban regions and the daily lives of residents. Urban region representation learning emerges as a crucial realm for fulfilling this task. Among deep learning approaches, graph neural networks (GNNs) have shown promise, given that city elements can be naturally represented as nodes with various connections between them as edges. However, many existing GNN approaches encounter challenges such as over-smoothing and limitations in capturing information from nodes in other regions, resulting in the loss of crucial urban information and a decline in region representation performance. To address these challenges, we leverage urban graph structure information and introduce a hierarchical graph pooling process called Coarsened Graph Attention Pooling (CGAP). CGAP features local attention units to create coarsened intermediate graphs and global features. Additionally, by incorporating urban region graphs and global features into a global attention layer, we harness relational information to enhance representation effectiveness. Furthermore, CGAP integrates region attributes such as Points of Interest (POIs) and inter-regional contexts like human mobility, enabling the exploitation of multi-modal urban data for more comprehensive representation learning. Experiments on three downstream tasks related to the UN Sustainable Development Goals validate the effectiveness of region representations learned by our approach. Experimental results and analyses demonstrate that CGAP excels in various socioeconomic prediction tasks compared to competitive baselines.

Yanxin Xi, Yu Liu, Zhicheng Liu, Sasu Tarkoma, Pan Hui, Yong Li

The Sustainable Development Goals (SDGs) aim to resolve societal challenges, such as eradicating poverty and improving the lives of vulnerable populations in impoverished areas. Those areas rely on road infrastructure construction to promote accessibility and economic development. Although publicly available data like OpenStreetMap is available to monitor road status, data completeness in impoverished areas is limited. Meanwhile, the development of deep learning techniques and satellite imagery shows excellent potential for earth monitoring. To tackle the challenge of road network assessment in impoverished areas, we develop a systematic road extraction framework combining an encoder-decoder architecture and morphological operations on satellite imagery, offering an integrated workflow for interdisciplinary researchers. Extensive experiments of road network extraction on real-world data in impoverished regions achieve a 42.7% enhancement in the F1-score over the baseline methods and reconstruct about 80% of the actual roads. We also propose a comprehensive road network dataset covering approximately 794,178 km2 area and 17.048 million people in 382 impoverished counties in China. The generated dataset is further utilized to conduct socioeconomic analysis in impoverished counties, showing that road network construction positively impacts regional economic development. The technical appendix, code, and generated dataset can be found at https://github.com/tsinghua-fib-lab/Road_network_extraction_impoverished_counties.

Jiasheng Wu, Xin Liu, Dawei Cheng, Yi Ouyang, Xian Wu, Yefeng Zheng

Financial fraud is one of the most significant social issues and has caused tremendous property losses. Graph neural networks (GNNs) have been applied to anti-fraud practices and achieved decent results. However, recent researches have discovered flaws in the robustness of fraud-detection models based on GNNs, enabling fraudsters to mislead them through attacks like data poisoning. In addition, most existing attack-defense models tend to study on ideal settings and lose information during truncation or filtering, which lowers their performances in complicated financial fraud cases. Therefore, in this paper, we propose a novel robust anti-fraud GNN model. In particular, we first design an attack algorithm tampering with both features and structures of graph data to simulate fraudsters' attacking behaviors in real-life complex fraud scenarios. Then we apply singular value decomposition to the graph and learn the decomposed matrices in a GNN model with specifically designed joint losses. This enables our model to learn the graph patterns in low-rank subspaces without losing too much detailed information and fit the graph structure to characteristics including class-homophily and sparsity to guarantee robustness. The proposed approach is experimented on real-world fraud datasets, which demonstrates its advantages in fraud detection and robustness compared with the state-of-the-art baselines.

Jiageng Wu, Xian Wu, Jie Yang

Clinical reasoning refers to the cognitive process that physicians employ in evaluating and managing patients. This process typically involves suggesting necessary examinations, diagnosing patients’ diseases, and selecting appropriate therapies, etc. Accurate clinical reasoning requires extensive medical knowledge and rich clinical experience, setting a high bar for physicians. This is particularly challenging in developing countries due to the overwhelming number of patients and limited physician resources, contributing significantly to global health inequity and necessitating automated clinical reasoning approaches. Recently, the emergence of large language models (LLMs) such as ChatGPT and GPT-4 have demonstrated their potential in clinical reasoning. However, these LLMs are prone to hallucination problems, and the reasoning process of LLMs may not align with the clinical decision pathways of physicians. In this study, we introduce a novel framework, In-Context Padding (ICP), to enhance LLMs reasoning with medical knowledge. Specifically, we infer critical clinical reasoning elements (referred to as knowledge seeds) and use these as anchors to guide the generation process of LLMs. Experiments on two clinical question datasets validate that ICP significantly improves the clinical reasoning ability of LLMs.