In the stereoscopic cage-rearing system, monitoring the individual egg production of laying ducks is essential for identifying low-yield individuals and optimizing breeding management. To overcome the high deployment cost of multi-sensor monitoring and the limited coverage of fixed cameras, this study proposes a mobile camera–based video monitoring method capable of automatically performing duck egg detection, cage QR code recognition, and egg–cage matching counting. Based on an improved YOLOv11 framework, a Lightweight Duck Egg and QR Code Detection model (LDEQ-OD) was developed. By integrating a Dual Detection Head (DH), a C3-DDF module, and a SENet attention mechanism, the model achieved notable improvements in small-object detection accuracy and real-time inference performance on edge devices. On the Jetson Nano platform, the model contained 1.4 M parameters and achieved an average inference time of 59.2 ms, with precision, recall, and mAP@0.5:0.95 reaching 99.6%, 99.3%, and 95.3%, respectively. The OC-SORT algorithm was employed for multi-object tracking to establish spatiotemporal associations between eggs and QR codes, while a Cascade Robust QR Code Decoding (CRQD) algorithm enhanced decoding accuracy under motion blur and uneven illumination. Compared with the traditional ZBar decoder, CRQD improved the overall Code Identification Rate from 72.7% to 99.3%, demonstrating significant robustness. Furthermore, a dynamic matching strategy based on the Minimum Aspect Ratio Deviation (MARD) was proposed to compensate for geometric distortion caused by camera tilt, achieving a Mean Absolute Error (MAE) of 0.017 eggs per cage and an Egg–Cage Matching Accuracy (ECMA) of 98.3%. Experiments under different motion speeds (0.14, 0.21, and 0.44 m/s) confirmed that the proposed matching algorithm maintained stable and reliable performance under complex geometric perspectives and varying camera movements. The system was deployed on the Jetson Nano platform at 10 frames per second and integrated with real-time data acquisition and visualization modules, enabling intelligent and real-time monitoring of individual egg production in cage-reared ducks to support precision breeding management.
Numerous recent studies have shown that Large Language Models (LLMs) are biased towards a Western and Anglo-centric worldview, which compromises their usefulness in non-Western cultural settings. However, "culture" is a complex, multifaceted topic, and its awareness, representation, and modeling in LLMs and LLM-based applications can be defined and measured in numerous ways. In this position paper, we ask what does it mean for an LLM to possess "cultural awareness", and through a thought experiment, which is an extension of the Octopus test proposed by Bender and Koller (2020), we argue that it is not cultural awareness or knowledge, rather meta-cultural competence, which is required of an LLM and LLM-based AI system that will make it useful across various, including completely unseen, cultures. We lay out the principles of meta-cultural competence AI systems, and discuss ways to measure and model those.
This paper examines the output of cultural items generated by Chat Generative PreTrained Transformer Pro in response to three structured prompts to translate three anthologies of African poetry. The first prompt was broad, the second focused on poetic structure, and the third prompt emphasized cultural specificity. To support this analysis, four comparative tables were created. The first table presents the results of the cultural items produced after the three prompts, the second categorizes these outputs based on Aixela framework of Proper nouns and Common expressions, the third table summarizes the cultural items generated by human translators, a custom translation engine, and a Large Language Model. The final table outlines the strategies employed by Chat Generative PreTrained Transformer Pro following the culture specific prompt. Compared to the outputs of cultural items from reference human translation and the custom translation engine in prior studies the findings indicate that the culture oriented prompts used with Chat Generative PreTrained Transformer Pro did not yield significant enhancements of cultural items during the translation of African poetry from English to French. Among the fifty four cultural items, the human translation produced thirty three cultural items in repetition, the custom translation engine generated Thirty eight cultural items in repetition while Chat Generative PreTrained Transformer Pro produced forty one cultural items in repetition. The untranslated cultural items revealed inconsistencies in Large language models approach to translating cultural items in African poetry from English to French.
Erfan Moosavi Monazzah, Vahid Rahimzadeh, Yadollah Yaghoobzadeh
et al.
Large language models predominantly reflect Western cultures, largely due to the dominance of English-centric training data. This imbalance presents a significant challenge, as LLMs are increasingly used across diverse contexts without adequate evaluation of their cultural competence in non-English languages, including Persian. To address this gap, we introduce PerCul, a carefully constructed dataset designed to assess the sensitivity of LLMs toward Persian culture. PerCul features story-based, multiple-choice questions that capture culturally nuanced scenarios. Unlike existing benchmarks, PerCul is curated with input from native Persian annotators to ensure authenticity and to prevent the use of translation as a shortcut. We evaluate several state-of-the-art multilingual and Persian-specific LLMs, establishing a foundation for future research in cross-cultural NLP evaluation. Our experiments demonstrate a 11.3% gap between best closed source model and layperson baseline while the gap increases to 21.3% by using the best open-weight model. You can access the dataset from here: https://huggingface.co/datasets/teias-ai/percul
Shuai Feng, Wei-Chuang Chan, Srishti Chouhan
et al.
The integration of large language models (LLMs) into global applications necessitates effective cultural alignment for meaningful and culturally-sensitive interactions. Current LLMs often lack the nuanced understanding required for diverse cultural contexts, and adapting them typically involves costly full fine-tuning. To address this, we introduce a novel soft prompt fine-tuning framework that enables efficient and modular cultural alignment. Our method utilizes vectorized prompt tuning to dynamically route queries to a committee of culturally specialized 'expert' LLM configurations, created by optimizing soft prompt embeddings without altering the base model's parameters. Extensive experiments demonstrate that our framework significantly enhances cultural sensitivity and adaptability, improving alignment scores from 0.208 to 0.820, offering a robust solution for culturally-aware LLM deployment. This research paves the way for subsequent investigations into enhanced cultural coverage and dynamic expert adaptation, crucial for realizing autonomous AI with deeply nuanced understanding in a globally interconnected world.
Serena Jinchen Xie, Shumenghui Zhai, Yanjing Liang
et al.
Large Language Model (LLM)-based conversational agents offer promising solutions for mental health support, but lack cultural responsiveness for diverse populations. This study evaluated the effectiveness of cultural prompting in improving cultural responsiveness and perceived empathy of LLM-generated therapeutic responses for Chinese American family caregivers. Using a randomized controlled experiment, we compared GPT-4o and Deepseek-V3 responses with and without cultural prompting. Thirty-six participants evaluated input-response pairs on cultural responsiveness (competence and relevance) and perceived empathy. Results showed that cultural prompting significantly enhanced GPT-4o's performance across all dimensions, with GPT-4o with cultural prompting being the most preferred, while improvements in DeepSeek-V3 responses were not significant. Mediation analysis revealed that cultural prompting improved empathy through improving cultural responsiveness. This study demonstrated that prompt-based techniques can effectively enhance the cultural responsiveness of LLM-generated therapeutic responses, highlighting the importance of cultural responsiveness in delivering empathetic AI-based therapeutic interventions to culturally and linguistically diverse populations.
Israel Abebe Azime, Tadesse Destaw Belay, Dietrich Klakow
et al.
Large language models (LLMs) have demonstrated significant capabilities in solving mathematical problems expressed in natural language. However, multilingual and culturally-grounded mathematical reasoning in low-resource languages lags behind English due to the scarcity of socio-cultural task datasets that reflect accurate native entities such as person names, organization names, and currencies. Existing multilingual benchmarks are predominantly produced via translation and typically retain English-centric entities, owing to the high cost associated with human annotater-based localization. Moreover, automated localization tools are limited, and hence, truly localized datasets remain scarce. To bridge this gap, we introduce a framework for LLM-driven cultural localization of math word problems that automatically constructs datasets with native names, organizations, and currencies from existing sources. We find that translated benchmarks can obscure true multilingual math ability under appropriate socio-cultural contexts. Through extensive experiments, we also show that our framework can help mitigate English-centric entity bias and improves robustness when native entities are introduced across various languages.
This paper introduces a Dual Evaluation Framework to comprehensively assess the multilingual capabilities of LLMs. By decomposing the evaluation along the dimensions of linguistic medium and cultural context, this framework enables a nuanced analysis of LLMs' ability to process questions within both native and cross-cultural contexts cross-lingually. Extensive evaluations are conducted on a wide range of models, revealing a notable "CulturalLinguistic Synergy" phenomenon, where models exhibit better performance when questions are culturally aligned with the language. This phenomenon is further explored through interpretability probing, which shows that a higher proportion of specific neurons are activated in a language's cultural context. This activation proportion could serve as a potential indicator for evaluating multilingual performance during model training. Our findings challenge the prevailing notion that LLMs, primarily trained on English data, perform uniformly across languages and highlight the necessity of culturally and linguistically model evaluations. Our code can be found at https://yingjiahao14. github.io/Dual-Evaluation/.
Sung Ho Yun, Md Arif-Ur Rahman, Soo-Bin Nam
et al.
Abstract The human norovirus (abbreviated as HuNV) is the most common agent responsible for acute viral gastroenteritis. Despite being recognized as a water-borne pathogenic virus for a long time, the cellular tropism of norovirus has not yet been clearly explained. The main reason is the lack of appropriate cell culture and animal model systems for HuNV infection. Murine norovirus (abbreviated as MNV) is often used as a proxy for human norovirus when trying to understand the expression profiles left behind by norovirus infection in a host. In the current study, the host response to MNV was examined using the macrophage Raw 264.7 in terms of the altered host proteomes. After MNV infection, host Raw 264.7 cell lysates were collected for proteome profiling at the time points of 0.5 hpi (early phase, control), 16 hpi (mid-phase), and 24 hpi (late phase). LC–MS analysis was employed for label-free shotgun proteomics on the host cell proteomes. The progression of MNV infection status was monitored using an immunofluorescence-conjugated noroviral capsid protein VP1 and a confocal microscope. The up-regulation of Ras GTPases such Rab5A and Rab6A was found to be implicated in norovirus gastroenteritis, as revealed by proteomic profiling. Consequently, the recognition of Ras-related proteins can lead to a better understanding of how noroviral infection affects the immune system of the host cell.
There is an urgent need to incorporate the perspectives of culturally diverse groups into AI developments. We present a novel conceptual framework for research that aims to expand, reimagine, and reground mainstream visions of AI using independent and interdependent cultural models of the self and the environment. Two survey studies support this framework and provide preliminary evidence that people apply their cultural models when imagining their ideal AI. Compared with European American respondents, Chinese respondents viewed it as less important to control AI and more important to connect with AI, and were more likely to prefer AI with capacities to influence. Reflecting both cultural models, findings from African American respondents resembled both European American and Chinese respondents. We discuss study limitations and future directions and highlight the need to develop culturally responsive and relevant AI to serve a broader segment of the world population.
Abdulla Alfalasi, Esrat Khan, Mohamed Alhashmi
et al.
A transformative approach to mental health therapy lies at the crossroads of cultural heritage and advanced technology. This paper introduces an innovative method that fuses machine learning techniques with traditional Emirati motifs, focusing on the United Arab Emirates (UAE). We utilize the Stable Diffusion XL (SDXL) model, enhanced with Low-Rank Adaptation (LoRA), to create culturally significant coloring templates featuring Al-Sadu weaving patterns. This novel approach leverages coloring therapy for its recognized stress-relieving benefits and embeds deep cultural resonance, making it a potent tool for therapeutic intervention and cultural preservation. Specifically targeting Generalized Anxiety Disorder (GAD), our method demonstrates significant potential in reducing associated symptoms. Additionally, the paper delves into the broader implications of color and music therapy, emphasizing the importance of culturally tailored content. The technical aspects of the SDXL model and its LoRA fine-tuning showcase its capability to generate high-quality, culturally specific images. This research stands at the forefront of integrating mental wellness practices with cultural heritage, providing a groundbreaking perspective on the synergy between technology, culture, and healthcare. In future work, we aim to employ biosignals to assess the level of engagement and effectiveness of color therapy. A key focus will be to examine the impact of the Emirati heritage Al Sadu art on Emirati individuals and compare their responses with those of other nationalities. This will provide deeper insights into the cultural specificity of therapeutic interventions and further the understanding of the unique interplay between cultural identity and mental health therapy.
Large language models (LLMs) closely interact with humans, and thus need an intimate understanding of the cultural values of human society. In this paper, we explore how open-source LLMs make judgments on diverse categories of cultural values across countries, and its relation to training methodology such as model sizes, training corpus, alignment, etc. Our analysis shows that LLMs can judge socio-cultural norms similar to humans but less so on social systems and progress. In addition, LLMs tend to judge cultural values biased toward Western culture, which can be improved with training on the multilingual corpus. We also find that increasing model size helps a better understanding of social values, but smaller models can be enhanced by using synthetic data. Our analysis reveals valuable insights into the design methodology of LLMs in connection with their understanding of cultural values.
Sk Injamamul Islam, Mohamed H. Hamad, Wanarit Jitsamai
et al.
Clinostomum species, a parasitic pathogen of freshwater fish, is widely distributed and infects various host species. Recently, the pathological effect due to Clinostomum metacercarial infection was described in aquaculture in Thailand; however, the global genetic diversity and population structure of this species have not been studied yet. Therefore, this study aimed to provide a detailed description of genetic diversity and population dynamics of the digenean Clinostomum isolated from Trichopodus pectoralis with globally recorded Clinostomum species. The species was characterized molecularly by analyzing 18S rDNA and inter-transcribed spacer biomarker genes (ITS1 and ITS2). A BLAST search discovered that the 18S rDNA and ITS sequence had a 100% sequence similarity with Clinostomum piscidium isolated from India and Thailand. A comprehensive analysis revealed the presence of 12 distinct haplotypes among the Clinostomum populations. This study suggests that distinct patterns of genetic variation were identified by analyzing molecular variance, pairwise Fst, and employing structure analysis. It was observed that a gradient of genetic variation exists within continents, characterized by higher levels within different groups and lower levels of genetic differentiation. Additionally, a notable presence of mixed haplotypes was observed. The results of neutrality testing suggest that there has been a significant expansion in the populations of Clinostomum in India, America, and Kenya. The discoveries from this study will provide a valuable contribution to comprehending the genetics and evolution of Clinostomum species. Furthermore, key findings will be essential in developing efficient management approaches to prevent and control this parasite.
Hiba S. Alnaemi, Tamara N. Dawood, Qais Th. Algwari
Objective: The efficiency of corona discharge (CD) for detoxification of aflatoxin B1 (AB1), ochra¬toxin A (OA), and fumonisin B1 (FMB1) from poultry feeds with its influences on feed components was investigated.
Materials and Methods: Feed samples were exposed to CD for six durations (10, 20, 30, 40, 50, and 60 min) at three distances (1.5, 2.5, and 3.5 cm). Mycotoxin levels were estimated by compet¬itive enzyme-linked immunosorbent assay, and findings were substantiated by high-performance liquid chromatography.
Results: AB1, OA, and FMB1 degradation percentages increased significantly (p < 0.05) with pro¬cessing times increment and distances reduction to reach values of 83.22%, 84.21%, and 84.76% at the first distance; 80.28%, 84.00%, and 84.12% at the second distance; and 68.30%, 71.74%, and 76.18% at the third distance, respectively, after 60 min of treatment. FMB1 reported the highest degradation level. Concerning CD impacts on feed composition, protein, fat, and moisture contents decreased significantly (p < 0.05). Carbohydrates and ash were not affected adversely. Depending on peroxide values estimation, fats were of good quality.
Conclusion: The CD effectiveness for AB1, OA, and FMB1 detox from poultry feeds with moderate impact on the quality of feed. [J Adv Vet Anim Res 2024; 11(4.000): 819-834]
Animal pose estimation is an important but under-explored task due to the lack of labeled data. In this paper, we tackle the task of animal pose estimation with scarce annotations, where only a small set of labeled data and unlabeled images are available. At the core of the solution to this problem setting is the use of the unlabeled data to compensate for the lack of well-labeled animal pose data. To this end, we propose the ScarceNet, a pseudo label-based approach to generate artificial labels for the unlabeled images. The pseudo labels, which are generated with a model trained with the small set of labeled images, are generally noisy and can hurt the performance when directly used for training. To solve this problem, we first use a small-loss trick to select reliable pseudo labels. Although effective, the selection process is improvident since numerous high-loss samples are left unused. We further propose to identify reusable samples from the high-loss samples based on an agreement check. Pseudo labels are re-generated to provide supervision for those reusable samples. Lastly, we introduce a student-teacher framework to enforce a consistency constraint since there are still samples that are neither reliable nor reusable. By combining the reliable pseudo label selection with the reusable sample re-labeling and the consistency constraint, we can make full use of the unlabeled data. We evaluate our approach on the challenging AP-10K dataset, where our approach outperforms existing semi-supervised approaches by a large margin. We also test on the TigDog dataset, where our approach can achieve better performance than domain adaptation based approaches when only very few annotations are available. Our code is available at the project website.
L. Cattaneo, V. Lopreiato, F. Piccioli-Cappelli
et al.
ABSTRACT: Dairy cows have to face several nutritional challenges during the transition period, and live yeast supplementation appears to be beneficial in modulating rumen activity. In this study, we evaluated the effects of live yeast supplementation on rumen function, milk production, and metabolic and inflammatory conditions. Ten Holstein multiparous cows received either live Saccharomyces cerevisiae (strain Sc47; SCY) supplementation from −21 to 21 d from calving (DFC) or a control diet without yeast supplementation. Feed intake, milk yield, and rumination time were monitored until 35 DFC, and rumen fluid, feces, milk, and blood samples were collected at different time points. Compared with the control diet, SCY had increased dry matter intake (16.7 vs. 19.1 ± 0.8 kg/d in wk 2 and 3) and rumination time postpartum (449 vs. 504 ± 19.9 min/d in wk 5). Milk yield tended to be greater in SCY (40.1 vs. 45.2 ± 1.7 kg/d in wk 5), protein content tended to be higher, and somatic cell count was lower. In rumen fluid, acetate molar proportion was higher and that of propionate lower at 21 DFC, resulting in increased acetate:propionate and (acetate + butyrate):propionate ratios. Cows in the SCY group had lower fecal dry matter but higher acetate and lower propionate proportions on total volatile fatty acids at 3 DFC. Plasma analysis revealed a lower degree of inflammation after calving in SCY (i.e., lower haptoglobin concentration at 1 and 3 DFC) and a likely better liver function, as suggested by the lower γ-glutamyl transferase, even though paraoxonase was lower at 28 DFC. Plasma IL-1β concentration tended to be higher in SCY, as well as Mg and P. Overall, SCY supplementation improved rumen and hindgut fermentation profiles, also resulting in higher dry matter intake and rumination time postpartum. Moreover, the postcalving inflammatory response was milder and liver function appeared to be better. Altogether, these effects also led to greater milk yield and reduced the risk of metabolic diseases.
ABSTRACT KIT protein is associated with the etiology of canine mast cell tumors (MCT); however, the expression patterns of KIT are highly variable. The aim of this study was to determine if KIT patterns are related with eosinophil count in MCT. Hematoxylin eosin and May Grünwald-Giemsa stain techniques were applied, histological grading and eosinophil counting were performed in 48 MCT samples. Immunohistochemical evaluation was performed with IL-5, VEGFr, and c-KIT antibodies. The percentage of immunolabeling with IL-5 and VEGFr was determined, and the samples incubated with c-KIT were graded according to the immunolabeling pattern. Comparison of the mean eosinophil count between the histological grades and the different KIT expression patterns demonstrated a significant difference between KIT pattern 1 and KIT pattern 3, KIT pattern 3 showed a higher mean of eosinophil count. There was no significant correlation between eosinophil count and KIT patterns (p = 0.2648). However, a positive correlation was observed between the KIT patterns and Patnaik and Kiupel grades (p = 0.0006 and p = 0.0267, respectively). There was no significant correlation between eosinophil count, IL-5, or VEGFr. Further studies should determine whether eosinophil counts are an independent predictor of clinical outcome or simply correlated with already known predictors.