Hasil untuk "Manners and customs (General)"

Chunlei Li, Jingyang Hou, Yilei Shi et al.

Medical report generation from imaging data remains a challenging task in clinical practice. While large language models (LLMs) show great promise in addressing this challenge, their effective integration with medical imaging data still deserves in-depth exploration. In this paper, we present MRG-LLM, a novel multimodal large language model (MLLM) that combines a frozen LLM with a learnable visual encoder and introduces a dynamic prompt customization mechanism. Our key innovation lies in generating instance-specific prompts tailored to individual medical images through conditional affine transformations derived from visual features. We propose two implementations: prompt-wise and promptbook-wise customization, enabling precise and targeted report generation. Extensive experiments on IU X-ray and MIMIC-CXR datasets demonstrate that MRG-LLM achieves state-of-the-art performance in medical report generation. Our code will be made publicly available.

Yimin Shi, Yang Fei, Shiqi Zhang et al.

In e-commerce, user representations are essential for various applications. Existing methods often use deep learning techniques to convert customer behaviors into implicit embeddings. However, these embeddings are difficult to understand and integrate with external knowledge, limiting the effectiveness of applications such as customer segmentation, search navigation, and product recommendations. To address this, our paper introduces the concept of the customer persona. Condensed from a customer's numerous purchasing histories, a customer persona provides a multi-faceted and human-readable characterization of specific purchase behaviors and preferences, such as Busy Parents or Bargain Hunters. This work then focuses on representing each customer by multiple personas from a predefined set, achieving readable and informative explicit user representations. To this end, we propose an effective and efficient solution GPLR. To ensure effectiveness, GPLR leverages pre-trained LLMs to infer personas for customers. To reduce overhead, GPLR applies LLM-based labeling to only a fraction of users and utilizes a random walk technique to predict personas for the remaining customers. We further propose RevAff, which provides an absolute error $ε$ guarantee while improving the time complexity of the exact solution by a factor of at least $O(\frac{ε\cdot|E|N}{|E|+N\log N})$, where $N$ represents the number of customers and products, and $E$ represents the interactions between them. We evaluate the performance of our persona-based representation in terms of accuracy and robustness for recommendation and customer segmentation tasks using three real-world e-commerce datasets. Most notably, we find that integrating customer persona representations improves the state-of-the-art graph convolution-based recommendation model by up to 12% in terms of NDCG@K and F1-Score@K.

Wooseok Song, Seunggyu Chang, Jaejun Yoo

While single-concept customization has been studied in 3D, multi-concept customization remains largely unexplored. To address this, we propose MultiDreamer3D that can generate coherent multi-concept 3D content in a divide-and-conquer manner. First, we generate 3D bounding boxes using an LLM-based layout controller. Next, a selective point cloud generator creates coarse point clouds for each concept. These point clouds are placed in the 3D bounding boxes and initialized into 3D Gaussian Splatting with concept labels, enabling precise identification of concept attributions in 2D projections. Finally, we refine 3D Gaussians via concept-aware interval score matching, guided by concept-aware diffusion. Our experimental results show that MultiDreamer3D not only ensures object presence and preserves the distinct identities of each concept but also successfully handles complex cases such as property change or interaction. To the best of our knowledge, we are the first to address the multi-concept customization in 3D.

Lu Liu, Huiyu Duan, Qiang Hu et al.

Artificial intelligence generative models exhibit remarkable capabilities in content creation, particularly in face image generation, customization, and restoration. However, current AI-generated faces (AIGFs) often fall short of human preferences due to unique distortions, unrealistic details, and unexpected identity shifts, underscoring the need for a comprehensive quality evaluation framework for AIGFs. To address this need, we introduce FaceQ, a large-scale, comprehensive database of AI-generated Face images with fine-grained Quality annotations reflecting human preferences. The FaceQ database comprises 12,255 images generated by 29 models across three tasks: (1) face generation, (2) face customization, and (3) face restoration. It includes 32,742 mean opinion scores (MOSs) from 180 annotators, assessed across multiple dimensions: quality, authenticity, identity (ID) fidelity, and text-image correspondence. Using the FaceQ database, we establish F-Bench, a benchmark for comparing and evaluating face generation, customization, and restoration models, highlighting strengths and weaknesses across various prompts and evaluation dimensions. Additionally, we assess the performance of existing image quality assessment (IQA), face quality assessment (FQA), AI-generated content image quality assessment (AIGCIQA), and preference evaluation metrics, manifesting that these standard metrics are relatively ineffective in evaluating authenticity, ID fidelity, and text-image correspondence. The FaceQ database will be publicly available upon publication.

Jianzong Wu, Chao Tang, Jingbo Wang et al.

Story visualization, the task of creating visual narratives from textual descriptions, has seen progress with text-to-image generation models. However, these models often lack effective control over character appearances and interactions, particularly in multi-character scenes. To address these limitations, we propose a new task: \textbf{customized manga generation} and introduce \textbf{DiffSensei}, an innovative framework specifically designed for generating manga with dynamic multi-character control. DiffSensei integrates a diffusion-based image generator with a multimodal large language model (MLLM) that acts as a text-compatible identity adapter. Our approach employs masked cross-attention to seamlessly incorporate character features, enabling precise layout control without direct pixel transfer. Additionally, the MLLM-based adapter adjusts character features to align with panel-specific text cues, allowing flexible adjustments in character expressions, poses, and actions. We also introduce \textbf{MangaZero}, a large-scale dataset tailored to this task, containing 43,264 manga pages and 427,147 annotated panels, supporting the visualization of varied character interactions and movements across sequential frames. Extensive experiments demonstrate that DiffSensei outperforms existing models, marking a significant advancement in manga generation by enabling text-adaptable character customization. The project page is https://jianzongwu.github.io/projects/diffsensei/.

Tereza Virginia de Almeida

A obscena Senhora D, de Hilda Hilst, tem como personagem central Hillé, uma viúva de sessenta anos que decide morar no vão da escada de sua casa, de onde fala com o marido e o pai já mortos, ouve vozes de dentro da parede e assusta a vizinhança com máscaras de papel. O artigo, dedicado a esta obra de Hilst, aborda o entrelaçamento entre velhice e loucura na definição de um espaço de exclusão a partir do qual a personagem experiencia uma singular relação com a linguagem. Trata-se não somente do constante contato com o tema da morte, mas também da experimentação em torno da palavra da qual deriva uma estética singular, uma poética, a poética da derrelição.

Hao Huang, Katherine R. Davis, H. Vincent Poor

The long-term resilient property of ecosystems has been quantified as ecological robustness (RECO) in terms of the energy transfer over food webs. The RECO of resilient ecosystems favors a balance of food webs' network efficiency and redundancy. By integrating RECO with power system constraints, the authors are able to optimize power systems' inherent resilience as ecosystems through network design and system operation. A previous model used on real power flows and aggregated redundant components for a rigorous mapping between ecosystems and power systems. However, the reactive power flows also determine power systems resilience; and the power components' redundancy is part of the global network redundancy. These characteristics should be considered for RECO-oriented evaluation and optimization for power systems. Thus, this paper extends the model for quantifying RECO in power systems using real, reactive, and apparent power flows with the consideration of redundant placement of generators. Recalling the performance of RECO-oriented optimal power flows under N-x contingencies, the analyses suggest reactive power flows and redundant components should be included for RECO to capture power systems' inherent resilience.

Pengyu Cheng, Jiawen Xie, Ke Bai et al.

Reward models (RMs) are essential for aligning large language models (LLMs) with human preferences to improve interaction quality. However, the real world is pluralistic, which leads to diversified human preferences with respect to different religions, politics, cultures, etc. Moreover, each individual can have their unique preferences on various topics. Neglecting the diversity of human preferences, current human feedback aligning methods only consider a general reward model, which is below satisfaction for customized or personalized application scenarios. To explore customized preference learning, we collect a domain-specific preference (DSP) dataset, which includes preferred responses for each given query from four practical domains. Besides, from the perspective of data efficiency, we propose a three-stage customized RM learning scheme, then empirically verify its effectiveness on both general preference datasets and our DSP set. Furthermore, we test multiple training and data strategies on the three learning stages. We find several ways to better preserve the general preferring ability while training the customized RMs, especially general preference enrichment, and customized preference imitation learning. The DSP dataset and code are available at https://github.com/Linear95/DSP.

James Seale Smith, Yen-Chang Hsu, Lingyu Zhang et al.

Recent works demonstrate a remarkable ability to customize text-to-image diffusion models while only providing a few example images. What happens if you try to customize such models using multiple, fine-grained concepts in a sequential (i.e., continual) manner? In our work, we show that recent state-of-the-art customization of text-to-image models suffer from catastrophic forgetting when new concepts arrive sequentially. Specifically, when adding a new concept, the ability to generate high quality images of past, similar concepts degrade. To circumvent this forgetting, we propose a new method, C-LoRA, composed of a continually self-regularized low-rank adaptation in cross attention layers of the popular Stable Diffusion model. Furthermore, we use customization prompts which do not include the word of the customized object (i.e., "person" for a human face dataset) and are initialized as completely random embeddings. Importantly, our method induces only marginal additional parameter costs and requires no storage of user data for replay. We show that C-LoRA not only outperforms several baselines for our proposed setting of text-to-image continual customization, which we refer to as Continual Diffusion, but that we achieve a new state-of-the-art in the well-established rehearsal-free continual learning setting for image classification. The high achieving performance of C-LoRA in two separate domains positions it as a compelling solution for a wide range of applications, and we believe it has significant potential for practical impact. Project page: https://jamessealesmith.github.io/continual-diffusion/

Anne Louise Dias

O subtítulo do romance Com armas sonolentas: um romance de formação aparece como um convite para que o leitor adentre a narrativa de Carola Saavedra já com uma chave de leitura em mãos. Com o espectro do bildungsroman em mente, ele busca ali uma história de formação nos moldes tradicionais, mas acaba encontrando a narrativa de três protagonistas, em fases diferentes de suas vidas; um certo desvio da tradição literária atribuída ao romance de formação na qual um jovem protagonista expõe seu processo de amadurecimento em direção à vida adulta. A quebra de expectativa é, assim, consciente. É objetivo do artigo compreender o que constituiria Com armas sonolentas um bildungsroman a partir de duas análises interdependentes no romance: a formação feminina das personagens protagonistas do texto e a narração de uma história que busca reconstruir o protagonismo de indivíduos outros na identidade da nação. A escrita passa a ser, portanto, recurso de uma reparação de traumas tanto em níveis individuais quanto nacionais.

Frederico Garcia Fernandes, Diego Giménez

A relação entre literatura, história e pensamento remonta às origens da cultura ocidental. Aristóteles, na Poética, afirma que o poeta e o historiador não diferem por um escrever em verso e o outro em prosa, mas pelo fato do primeiro narrar o que poderia acontecer e o segundo o que aconteceu. Assim, para o filósofo do Liceu, a poesia é mais filosófica do que a história. Aquela procura o universal, esta, o particular. Essas afirmações foram contestadas ao longo do tempo e no decorrer de teorias e estéticas que questionaram os limites e as possibilidades quer da narração, quer do pensamento, quer da própria história. Como a narrativa lusófona enfrenta as questões levantadas por Georg Lukács, Walter Benjamin ou Adorno, se uma totalidade verossímil não é mais dada às formas?O número 36 da revista Veredas convida os leitores a pensar como se entrelaçam literatura, pensamento e história nos confins do narrável, mediante uma série de textos que interrogam a própria possibilidade da literatura.

Anthony Melaragno, William Casey

Ransomware has been an ongoing issue since the early 1990s. In recent times ransomware has spread from traditional computational resources to cyber-physical systems and industrial controls. We devised a series of experiments in which virtual instances are infected with ransomware. We instrumented the instances and collected resource utilization data across a variety of metrics (CPU, Memory, Disk Utility). We design a change point detection and learning method for identifying ransomware execution. Finally we evaluate and demonstrate its ability to detect ransomware efficiently in a timely manner when trained on a minimal set of samples. Our results represent a step forward for defense, and we conclude with further remarks for the path forward.

Pin-Jung Chen, Sahil Bhatnagar, Sagar Goyal et al.

Tasks critical to enterprise profitability, such as customer churn prediction, fraudulent account detection or customer lifetime value estimation, are often tackled by models trained on features engineered from customer data in tabular format. Application-specific feature engineering adds development, operationalization and maintenance costs over time. Recent advances in representation learning present an opportunity to simplify and generalize feature engineering across applications. When applying these advancements to tabular data researchers deal with data heterogeneity, variations in customer engagement history or the sheer volume of enterprise datasets. In this paper, we propose a novel approach to encode tabular data containing customer transactions, purchase history and other interactions into a generic representation of a customer's association with the business. We then evaluate these embeddings as features to train multiple models spanning a variety of applications. CASPR, Customer Activity Sequence-based Prediction and Representation, applies Transformer architecture to encode activity sequences to improve model performance and avoid bespoke feature engineering across applications. Our experiments at scale validate CASPR for both small and large enterprise applications.

Gunnar Kudrjavets, Jeff Thomas, Aditya Kumar et al.

Using custom memory allocators is an efficient performance optimization technique. However, dependency on a custom allocator can introduce several maintenance-related issues. We present lessons learned from the industry and provide critical guidance for using custom memory allocators and enumerate various challenges associated with integrating them. These recommendations are based on years of experience incorporating custom allocators into different industrial software projects.

Joseane Maytê Sousa Santos Sousa

O presente artigo visa discutir de forma crítica e pessoal a produção de literatura dentro das escolas, agente letrador por excelência. Inicialmente é feita uma apresentação acerca dos conceitos de letramentos como prática social, que promove a inclusão social. Em seguida, são apontados diversos exemplos e juntamente com eles questionamentos acerca da negação de direito à literatura dentro das unidades escolares públicas, sobretudo à população negra que dela faz parte, o que contraria à proposta humanizadora do trabalho com a leitura de literatura. Por fim, é apresentada a produção da literatura pelos estudantes como potência de existir no mundo e dizer-se, portanto, enquanto resistência, o que dialoga com o tema do VI Encontro de Leitura e Literatura da UNEB (Universidade Estadual da Bahia), cujo tema central foi Direitos Humanos, Leitura e Literatura: Criar, Existir e Resistir.

Laila Loudiki, Mustapha Kchikech, El Hassan Essaky

Due to their broad application to different fields of theory and practice, generalized Petersen graphs $GPG(n,s)$ have been extensively investigated. Despite the regularity of generalized Petersen graphs, determining an exact formula for the diameter is still a difficult problem. In their paper, Beenker and Van Lint have proved that if the circulant graph $C_n(1,s)$ has diameter $d$, then $GPG(n,s)$ has diameter at least $d+1$ and at most $d+2$. In this paper, we provide necessary and sufficient conditions so that the diameter of $GPG(n,s)$ is equal to $d+1,$ and sufficient conditions so that the diameter of $GPG(n,s)$ is equal to $d+2.$ Afterwards, we give exact values for the diameter of $GPG(n,s)$ for almost all cases of $n$ and $s.$ Furthermore, we show that there exists an algorithm computing the diameter of generalized Petersen graphs with running time $O$(log$n$).

Akshay Agarwal, Shashank Maiya, Sonu Aggarwal

Customer service is a setting that calls for empathy in live human agent responses. Recent advances have demonstrated how open-domain chatbots can be trained to demonstrate empathy when responding to live human utterances. We show that a blended skills chatbot model that responds to customer queries is more likely to resemble actual human agent response if it is trained to recognize emotion and exhibit appropriate empathy, than a model without such training. For our analysis, we leverage a Twitter customer service dataset containing several million customer<->agent dialog examples in customer service contexts from 20 well-known brands.

Yang Liu, Yifei Sun, Vincent Gao

E-commerce stores collect customer feedback to let sellers learn about customer concerns and enhance customer order experience. Because customer feedback often contains redundant information, a concise summary of the feedback can be generated to help sellers better understand the issues causing customer dissatisfaction. Previous state-of-the-art abstractive text summarization models make two major types of factual errors when producing summaries from customer feedback, which are wrong entity detection (WED) and incorrect product-defect description (IPD). In this work, we introduce a set of methods to enhance the factual consistency of abstractive summarization on customer feedback. We augment the training data with artificially corrupted summaries, and use them as counterparts of the target summaries. We add a contrastive loss term into the training objective so that the model learns to avoid certain factual errors. Evaluation results show that a large portion of WED and IPD errors are alleviated for BART and T5. Furthermore, our approaches do not depend on the structure of the summarization model and thus are generalizable to any abstractive summarization systems.

Fauziah Ismi Desiana, Reiza D. Dienaputra

Akulturasi budaya Jepang dan Sunda dalam bingkai tata rias Sunda Siger membuktikan bahwa tata rias tradisional dapat dikemas modern dalam balutan teknik Igari Look. Penelitian ini menggunakan metode kualitatif dengan pendekatan deskriptif. Pengumpulan data didapatkan dari wawancara dengan make-up artist yang menggunakan teknik make-up Igari Look dalam tata rias Sunda Siger dan aktif mengunggah hasil tata riasnya dalam sebuah portofolio di Instagram. Akulturasi kebudayaan Sunda dan Jepang dalam tata rias Sunda Siger merupakan bukti bahwa masyarakat Sunda terbuka dengan situasi multikultural. Keberadaan Igari Look dalam bingkai tata rias Sunda Siger pada hakikatnya bukan bertujuan untuk memarjinalkan makna filosofis dan historis dalam setiap unsur tata rias Sunda Siger, namun proses inovasi dari make-up artist ini perlu dimaknai sebagai sumbangsih untuk menghidupkan kembali tata rias tradisional agar lebih diminati oleh kaum muda.   Acculturation of Japanese and Sundanese culture in frame Sunda Siger cosmetology proves that traditional cosmetology can be filled with modern dressing in the Igari Look technique. This study uses qualitative methods using descriptive. Data collection was obtained from interviews with make-up artists who used the Igari Look make-up technique in the Sunda Siger makeup and actively uploaded the makeup results in a portfolio on Instagram. Acculturation of Sundanese and Japanese culture in Sundanese Siger makeup is proof that Sundanese society is open with multiculturalism. The existence of Igari Look in the Sunda Siger makeup frame in essence is not an agreement to marginalize philosophical and historical meanings in any Sundanese Siger makeup, the innovation process of this make-up artist needs to be interpreted as cleft of young people.

Irina Basieva, Víctor H. Cervantes, Ehtibar N. Dzhafarov et al.

In quantum physics there are well-known situations when measurements of the same property in different contexts (under different conditions) have the same probability distribution, but cannot be represented by one and the same random variable. Such systems of random variables are called contextual. More generally, true contextuality is observed when different contexts force measurements of the same property (in psychology, responses to the same question) to be more dissimilar random variables than warranted by the difference of their distributions. The difference in distributions is itself a form of context-dependence, but of another nature: it is attributable to direct causal influences exerted by contexts upon the random variables. The Contextuality-by-Default (CbD) theory allows one to separate true contextuality from direct influences in the overall context-dependence. The CbD analysis of numerous previous attempts to demonstrate contextuality in human judgments shows that all context-dependence in them can be accounted for by direct influences, with no true contextuality present. However, contextual systems in human behavior can be found. In this paper we present a series of crowdsourcing experiments that exhibit true contextuality in simple decision making.}{The design of these experiments is an elaboration of one introduced in the "Snow Queen" experiment (Decision 5, 193-204, 2018), where contextuality was for the first time demonstrated unequivocally.