Hasil untuk "Animal culture"

Lucas Martini, Alexander Lappe, Anna Bognár et al.

The recognition of dynamic and social behavior in animals is fundamental for advancing ethology, ecology, medicine and neuroscience. Recent progress in deep learning has enabled automated behavior recognition from video, yet an accurate reconstruction of the three-dimensional (3D) pose and shape has not been integrated into this process. Especially for non-human primates, mesh-based tracking efforts lag behind those for other species, leaving pose descriptions restricted to sparse keypoints that are unable to fully capture the richness of action dynamics. To address this gap, we introduce the $\textbf{Big Ma}$ca$\textbf{Q}$ue 3D Motion and Animation Dataset ($\texttt{BigMaQ}$), a large-scale dataset comprising more than 750 scenes of interacting rhesus macaques with detailed 3D pose descriptions. Extending previous surface-based animal tracking methods, we construct subject-specific textured avatars by adapting a high-quality macaque template mesh to individual monkeys. This allows us to provide pose descriptions that are more accurate than previous state-of-the-art surface-based animal tracking methods. From the original dataset, we derive BigMaQ500, an action recognition benchmark that links surface-based pose vectors to single frames across multiple individual monkeys. By pairing features extracted from established image and video encoders with and without our pose descriptors, we demonstrate substantial improvements in mean average precision (mAP) when pose information is included. With these contributions, $\texttt{BigMaQ}$ establishes the first dataset that both integrates dynamic 3D pose-shape representations into the learning task of animal action recognition and provides a rich resource to advance the study of visual appearance, posture, and social interaction in non-human primates. The code and data are publicly available at https://martinivis.github.io/BigMaQ/ .

Weihan Li, Jingyang Ke, Yule Wang et al.

Understanding animal behavior from video is essential for neuroscience research. Modern laboratories typically collect two complementary data streams: skeletal keypoints from pose estimation tools and raw video recordings. Keypoint-based methods are efficient but suffer from geometric ambiguity, environmental blindness, and sensitivity to occlusions. Video-based methods capture rich context but require processing every frame, making them impractical for the hundreds of hours of recordings that modern experiments produce. We introduce LookAgain, a multimodal framework that combines the efficiency of keypoints with the representational power of video through on-demand visual grounding. During training, LookAgain uses dense visual features to pretrain a motion encoder and to train a gating module that learns which frames require visual context. During inference, this gating module activates visual processing only when keypoint signals are ambiguous, while maintaining performance comparable to using all frames. Experiments on single-animal and multi-animal benchmarks show that LookAgain achieves strong performance with significantly reduced computational cost, enabling high-quality behavior analysis on long-duration recordings.

Seungho Cho, Changgeon Ko, Eui Jun Hwang et al.

Large language models (LLMs) are increasingly used across diverse cultural contexts, making accurate cultural understanding essential. Prior evaluations have mostly focused on output-level performance, obscuring the factors that drive differences in responses, while studies using circuit analysis have covered few languages and rarely focused on culture. In this work, we trace LLMs' internal cultural understanding mechanisms by measuring activation path overlaps when answering semantically equivalent questions under two conditions: varying the target country while fixing the question language, and varying the question language while fixing the country. We also use same-language country pairs to disentangle language from cultural aspects. Results show that internal paths overlap more for same-language, cross-country questions than for cross-language, same-country questions, indicating strong language-specific patterns. Notably, the South Korea-North Korea pair exhibits low overlap and high variability, showing that linguistic similarity does not guarantee aligned internal representation.

Aditya Tomar, Nihar Ranjan Sahoo, Ashish Mittal et al.

Although mathematics is often considered culturally neutral, the way mathematical problems are presented can carry implicit cultural context. Existing benchmarks like GSM8K are predominantly rooted in Western norms, including names, currencies, and everyday scenarios. In this work, we create culturally adapted variants of the GSM8K test set for five regions Africa, India, China, Korea, and Japan using prompt-based transformations followed by manual verification. We evaluate six large language models (LLMs), ranging from 8B to 72B parameters, across five prompting strategies to assess their robustness to cultural variation in math problem presentation. Our findings reveal a consistent performance gap: models perform best on the original US-centric dataset and comparatively worse on culturally adapted versions. However, models with reasoning capabilities are more resilient to these shifts, suggesting that deeper reasoning helps bridge cultural presentation gaps in mathematical tasks

Chris M. Harrison, James W. Trayford, Arron George et al.

Sonification can be part of educational resources that can be accessible to those who prefer, or require, non-visual learning methods. Furthermore, sonification can contribute to an engaging multi-sensory learning experience, which are known to benefit general learners. Whilst some sonification can be relatively agnostic to musical culture, many sonifications are subject to culturally influenced choices, such as the chosen harmonies, rhythmic structures, and instrumentation. This is important when considering how universally inclusive and relatable sonification-based educational resources will be. Here we present a case study of translating a sonification-based educational show about the Solar System, that was originally designed with influences from Euro-American (Western-classical) music, to be more culturally relevant to the Caribbean region. We describe the motivation, approach, some of the challenges, and the initial feedback of the resulting output of the project. Finally, we provide reflections on the importance of further work exploring how educational sonifications can transcend international borders and musical cultures.

Emil Mededovic, Valdy Laurentius, Yuli Wu et al.

We analyze the capabilities of foundation models addressing the tedious task of generating annotations for animal tracking. Annotating a large amount of data is vital and can be a make-or-break factor for the robustness of a tracking model. Robustness is particularly crucial in animal tracking, as accurate tracking over long time horizons is essential for capturing the behavior of animals. However, generating additional annotations using foundation models can be counterproductive, as the quality of the annotations is just as important. Poorly annotated data can introduce noise and inaccuracies, ultimately compromising the performance and accuracy of the trained model. Over-reliance on automated annotations without ensuring precision can lead to diminished results, making careful oversight and quality control essential in the annotation process. Ultimately, we demonstrate that a thoughtful combination of automated annotations and manually annotated data is a valuable strategy, yielding an IDF1 score of 80.8 against blind usage of SAM2 video with an IDF1 score of 65.6.

Priyanka Dey, Yugal Khanter, Aayush Bothra et al.

As LLMs become central to interactive applications, ranging from tutoring to mental health, the ability to express personality in culturally appropriate ways is increasingly important. While recent works have explored personality evaluation of LLMs, they largely overlook the interplay between culture and personality. To address this, we introduce CulturalPersonas, the first large-scale benchmark with human validation for evaluating LLMs' personality expression in culturally grounded, behaviorally rich contexts. Our dataset spans 3,000 scenario-based questions across six diverse countries, designed to elicit personality through everyday scenarios rooted in local values. We evaluate three LLMs, using both multiple-choice and open-ended response formats. Our results show that CulturalPersonas improves alignment with country-specific human personality distributions (over a 20% reduction in Wasserstein distance across models and countries) and elicits more expressive, culturally coherent outputs compared to existing benchmarks. CulturalPersonas surfaces meaningful modulated trait outputs in response to culturally grounded prompts, offering new directions for aligning LLMs to global norms of behavior. By bridging personality expression and cultural nuance, we envision that CulturalPersonas will pave the way for more socially intelligent and globally adaptive LLMs.

Eric Neiva, Hervé Turlier

This work presents a novel unfitted finite element framework to simulate coupled surface-bulk problems in time-dependent domains, focusing on fluid-fluid interactions in animal cells between the actomyosin cortex and the cytoplasm. The cortex, a thin layer beneath the plasma membrane, provides structural integrity and drives shape changes by generating surface contractile forces akin to tension. Cortical contractions generate Marangoni-like surface flows and induce intracellular cytoplasmic flows that are essential for processes such as cell division, migration, and polarization, particularly in large animal cells. Despite its importance, the spatiotemporal regulation of cortex-cytoplasm interactions remains poorly understood and computational modelling can be very challenging because surface-bulk dynamics often lead to large cell deformations. To address these challenges, we propose a sharp-interface framework that uniquely combines the trace finite element method for surface flows with the aggregated finite element method for bulk flows. This approach enables accurate and stable simulations on fixed Cartesian grids without remeshing. The model also incorporates mechanochemical feedback through the surface transport of a molecular regulator of active tension. We solve the resulting mixed-dimensional system on a fixed Cartesian grid using a level-set-based method to track the evolving surface. Numerical experiments validate the accuracy and stability of the method, capturing phenomena such as self-organised pattern formation, curvature-driven relaxation, and cell cleavage. This novel framework offers a powerful and extendable tool for investigating increasingly complex morphogenetic processes in animal cells.

Yanli Chen, Yan Wang, Weike Shaoyong et al.

Abstract Background Compelling evidence has established a strong link between the gut microbiota and host reproductive health. However, the specific regulatory roles of individual bacterial species on reproductive performance are not well-understood. In the present study, Jinhua sows with varying reproductive performances under the same diet and management conditions were selected to explore potential mechanisms on the intricate relationship between the gut microbiome and host reproductive performance using 16S rRNA sequencing, metagenomics and serum metabolomics. Results Our findings revealed that the KEGG pathways for base excision repair and DNA replication were enriched, along with gene-level enhancements in spore formation, in sows with higher reproductive performance, indicating that the gut microbiome experiences stress. Further analysis showed a positive correlation between these changes and litter size, indicating that the host acts as a stressor, reshaping the microbiome. This adaptation allows the intestinal microbes in sows with high reproductive performance to enrich specific serotonin-related bacteria, such as Oxalobacter formigenes, Ruminococcus sp. CAG 382, Clostridium leptum, and Clostridium botulinum. Subsequently, the enriched microbiota may promote host serotonin production, which is positively correlated with reproductive performance in our study, known to regulate follicle survival and oocyte maturation. Conclusion Our study provides a theoretical basis for understanding the interactions between gut microbes and the host. It highlights new insights into reassembling gut microbiota in sows with higher litter sizes and the role of serotonin-related microbiota and serotonin in fertility. Graphical Abstract

Per Wallgren, Sven-Erik Johansson, Mikael Kirk et al.

Abstract Background Adult sows spend more than 50% of their time in units for gestating sows. Consequently, the functionality of these facilities is important for their well-being. This project aimed to depict the well-being of loose housed gestating sows in uninsulated buildings by documenting behaviours, aggressions, stereotypical behaviours and treatments/cullings of sows. Sows were fed individually with an animal-adapted transponder technology with four eating cubicles that allowed individual feeding adjustments. Between meals, the sows could choose from occupying themselves in a tent with deep litter straw or a barn with sawdust and chopped straw. Results Sows weaned 13.0 ± 0.7 piglets per litter at a mean age of 31 days with a mean weight of 11.2 ± 0.9 kg. Everyday monitoring and handling of 150–160 gestating sows demanded 32 ± 12 minutes per day of the staff. Medical treatments (9%) and cullings (1%) of non-lactating sows were rare. During gestation, most of the sows preferred to stay in the tent where they mainly rested, but daytime there were always sows in the barn where they were more active. Air temperatures and humidity remained comfortable throughout the year but differed between seasons. During spring, summer and autumn, queuing in front of the eating cubicles were rare and few aggressions were recorded. The activity of the sows was lowest during summer when sows also rested individually. Sows were most active during winter when they rested and moved in groups. Consequently, queuing and interactions in front of the feeding cubicles increased. However, most of these interactions were directed to the side or rear of another sow, and not against the head. Sows ate hierarchically, old sows ate first during the day and gilts last. No stereotypic behaviour was recorded. Conclusions Providing sows with varied residential options between meals and individually adjusted feeding during the gestating period ensured motion and well-being without reducing productivity and with low incidences of aggressions and medical treatments/cullings. The study also confirmed that gestating sows can be housed in uninsulated buildings during winter, if they are shielded from wind and can guard themselves from chill, e.g. by bivouacking themselves in deep litter straw.

Tuan-Phong Nguyen, Simon Razniewski, Gerhard Weikum

Despite recent progress, large language models (LLMs) still face the challenge of appropriately reacting to the intricacies of social and cultural conventions. This paper presents MANGO, a methodology for distilling high-accuracy, high-recall assertions of cultural knowledge. We judiciously and iteratively prompt LLMs for this purpose from two entry points, concepts and cultures. Outputs are consolidated via clustering and generative summarization. Running the MANGO method with GPT-3.5 as underlying LLM yields 167K high-accuracy assertions for 30K concepts and 11K cultures, surpassing prior resources by a large margin in quality and size. In an extrinsic evaluation for intercultural dialogues, we explore augmenting dialogue systems with cultural knowledge assertions. Notably, despite LLMs inherently possessing cultural knowledge, we find that adding knowledge from MANGO improves the overall quality, specificity, and cultural sensitivity of dialogue responses, as judged by human annotators. Data and code are available for download.

S.J.R. Woodward, J.P. Edwards, K.J. Verhoek et al.

Heat stress events in dairy cows are associated with behavioral and physiological changes such as seeking shade, increased respiration rate and body temperature, reduced milk production, and psychological distress. Knowledge of the relationship between weather and animal responses to heat stress enables automated alerts using forecast weather, aiding early provision of shade or other mitigation practices. While numerous heat stress indices for cattle have been developed, these have limitations for cows exposed to wind and solar radiation (i.e., predominantly grazing outdoors or managed on pasture). To develop a predictive model for heat stress events in pasture-based dairy systems, rumen temperature data from smaXtec (smaXtec animal care GmbH, Graz, Austria) rumen boluses in 443 cows on 3 dairy farms in Northland, New Zealand, were used to identify heat stress events and these were matched with automated weather station data collected on or near the farm. Heat stress rate (HSR) was defined as the percentage of cows within an age-breed group having a rumen temperature greater than 3 standard deviations above an individual cow's mean and heat stress events were defined as HSR >25%. Single and multiple linear regression models, including published heat stress indices, were generally able to predict a high proportion of heat stress events (sensitivity 34%–68%), but were insufficiently discriminating, predicting also a high number of false positives (precision only 9%–27%). A machine learning algorithm, cubist, was the best performing model, predicting 79% of heat stress events with a precision of 52% for this dataset. Our proof-of-concept study demonstrates the potential of this approach, using climate data to predict and forecast heat stress events in pasture-based dairy systems. Further work should test the cubist model using independent data, refine dataset construction, investigate the value of including known animal variables such as cow age or breed, and incorporate other measures of heat stress such as respiration rate.

Ahmet Haydar Ornek, Deniz Sen, Esmanur Civil

The main aim of this study is to demonstrate how innovative view synthesis and 3D reconstruction techniques can be used to create models of endangered species using monocular RGB images. To achieve this, we employed SyncDreamer to produce unique perspectives and NeuS and NeRF to reconstruct 3D representations. We chose four different animals, including the oriental stork, frog, dragonfly, and tiger, as our subjects for this study. Our results show that the combination of SyncDreamer, NeRF, and NeuS techniques can successfully create 3D models of endangered animals. However, we also observed that NeuS produced blurry images, while NeRF generated sharper but noisier images. This study highlights the potential of modeling endangered animals and offers a new direction for future research in this field. By showcasing the effectiveness of these advanced techniques, we hope to encourage further exploration and development of techniques for preserving and studying endangered species.

Katharina Dieck, Herwig Grimm

Despite the challenges the pandemic presented for university teaching, it opened up opportunities to set up and explore digital teaching formats like never before. This paper presents a case study of teaching introductory animal ethics in a digital format with flipped-classroom methods. The Interactive Literature Lecturing Format (ILLF) was designed along the following criteria: 1. Conformity with students’ varying educational needs; 2. Consistent high level of interaction; 3. Maximum transparency in an application-oriented exam; 4. No further contribution to the workload of the teaching staff; 5. Flexibility regarding online or on-site conversions. Rather than provide the students with input in lecture sessions, the ILLF presents students with selected literature and a list of structured questions. This literature questionnaire serves as the main didactic element that guides the knowledge transfer, the structure of the sessions and the exam. This paper reviews the outcome of the redesigning process and the steps we took to implement it. To discuss the overall quality of the format from a student’s perspective, the data from the systematically conducted students’ evaluation (<i>n</i> = 65) are interpreted using quantitative and qualitative methods. Bringing these results together with the perspective of the teaching staff, the following question is discussed: did the ILLF meet these criteria? This case study explores the potential and limits of flipped-classroom methods for applied ethics teaching in a university setting.

Pierrick Pochelu, Clara Erard, Philippe Cordier et al.

Camera traps have revolutionized the animal research of many species that were previously nearly impossible to observe due to their habitat or behavior. They are cameras generally fixed to a tree that take a short sequence of images when triggered. Deep learning has the potential to overcome the workload to automate image classification according to taxon or empty images. However, a standard deep neural network classifier fails because animals often represent a small portion of the high-definition images. That is why we propose a workflow named Weakly Object Detection Faster-RCNN+FPN which suits this challenge. The model is weakly supervised because it requires only the animal taxon label per image but doesn't require any manual bounding box annotations. First, it automatically performs the weakly-supervised bounding box annotation using the motion from multiple frames. Then, it trains a Faster-RCNN+FPN model using this weak supervision. Experimental results have been obtained with two datasets from a Papua New Guinea and Missouri biodiversity monitoring campaign, then on an easily reproducible testbed.

Hugh. R. Duddy, Mike. J. Schoonover, Brent. A. Hague

Abstract Background Tendon and ligament injuries are significant causes of loss of use and early retirement in performance horses. Amniotic fluid and tissue are excellent sources of growth factors and cytokines important in tendon and ligament healing. Thus, an equine-origin liquid amnion allograft (ELAA) may be beneficial in the treatment of equine tendonitis and desmitis. Objectives of this study were to report the outcome achieved (i.e. ability to return to work) for horses diagnosed with tendonitis or desmitis lesions treated with local injection of ELAA and to compare these outcomes to those reported for other regenerative medicine modalities. Methods A prospective, multi-center, non-blinded clinical trial was conducted. Equine veterinarians at 14 sites were selected to participate in the data collection for the trial. Criterion for inclusion was a horse presenting with lameness which was attributed to tendonitis or desmitis by diagnostic anesthesia and/or imaging. These horses were subsequently treated by local injection of the lesion with ELAA by the attending veterinarian. Standardized questionnaires describing each horse’s signalment, discipline, ability to return to work, and any adverse events were completed and submitted by the attending veterinarian following a minimum of six months of follow-up. The current literature was reviewed to identify clinical studies reporting outcomes of equine tendonitis/desmitis lesions treated with other regenerative therapies. Contingency table analyses were performed comparing outcomes. Results Questionnaires for 100 horses with 128 tendonitis and desmitis lesions met the inclusion criteria. Of these, 72 horses with 94 lesions returned to or exceeded their original level of work, 10 horses with 13 lesions returned to work but could not perform to previous standards, and 18 horses with 20 lesions did not return to work as a result of the injury. No differences were observed when outcome of horses treated with ELAA were compared to those of similar studies using other regenerative therapies. Conclusions Treatment of tendonitis and desmitis lesions by local injection of ELAA resulted in similar outcomes for horses returning to previous level of performance as other regenerative modalities such as mesenchymal stem cells, platelet-rich plasma, and autologous conditioned serum; however, blinded placebo-controlled studies are indicated.

Garrett Eickelberg, Yuan Luo, L. Nelson Sanchez-Pinto

Microbiology culture reports contain critical information for important clinical and public health applications. However, microbiology reports often have complex, semi-structured, free-text data that present a barrier for secondary use. Here we present the development and validation of an open-source package designed to ingest free-text microbiology reports, determine whether the culture is positive, and return a list of SNOMED-CT mapped bacteria. Our rule-based natural language processing algorithm was developed using microbiology reports from two different electronic health record systems in a large healthcare organization, and then externally validated on the reports of two other institutions with manually-extracted results as a benchmark. Our algorithm achieved F-1 scores >0.95 on all classification tasks across both validation sets. Our concept extraction Python package, MicrobEx, is designed to be reused and adapted to individual institutions as an upstream process for other clinical applications, such as machine learning studies, clinical decision support, and disease surveillance systems.

Giovanni Niero, Tania Bobbo, Simone Callegaro et al.

Extensive summer grazing is a dairy herd management practice frequently adopted in mountainous areas. Nowadays, this activity is threatened by its high labour demand, but it is fundamental for environmental, touristic and economic implications, as well as for the preservation of social and cultural traditions. Scarce information on the effects of such low-input farming systems on cattle health is available. Therefore, the present case study aimed at investigating how grazing may affect the health status of dairy cows by using milk traits routinely available from the national milk recording scheme. The research involved a dairy herd of 52 Simmental and 19 Holstein × Simmental crossbred cows. The herd had access to the pasture according to a rotational grazing scheme from late spring up to the end of summer. A total of 616 test day records collected immediately before and during the grazing season were used. Individual milk yield was registered during the milking procedure. Milk samples were analysed for composition (fat, protein, casein and lactose contents) and health-related milk indicators (electrical conductivity, urea and β-hydroxybutyrate) using mid-infrared spectroscopy. Somatic cell count (SCC) and differential SCC were also determined. Data were analysed with a linear mixed model, which included the fixed effects of the period of sampling, cow breed, stage of lactation and parity, and the random effects of cow nested within breed and the residual. The transition from barn farming to pasture had a negative effect on milk yield, together with a small deterioration of fat and protein percentages. Health-related milk indicators showed a minor deterioration of the fat to protein ratio, differential SCC and electrical conductivity, particularly towards the end of the grazing season, whereas the somatic cell score and β-hydroxybutyrate were relatively constant. Overall, the study showed that, when properly managed, pasture grazing does not have detrimental effects on dairy cows in terms of udder health and efficiency. Therefore, the proper management of cows on pasture can be a valuable solution to preserve the economic, social and environmental sustainability of small dairy farms in the alpine regions, without impairing cows’ health.

Shteryo Nozharov

The publication is one of the first studies of its kind, devoted to the economic dimension of crimes against cultural and archaeological heritage. Lack of research in this area is largely due to irregular global prevalence vague definition of economic value of the damage these crimes cause to the society at national and global level, to present and future generations. The author uses classical models of Becker and Freeman, by modifying and complementing them with the tools of economics of culture based on the values of non-use. The model tries to determine the opportunity costs of this type of crime in several scenarios and based on this to determine the extent of their limitation at an affordable cost to society and raising public benefits of conservation of World and National Heritage.

Sergei Nechaev, Michael Tamm

We provide combinatorial arguments based on a two-dimensional extension of a locally-free semigroup allowing us to compute the growth rate, $Λ$, of the partition function $Z_N=N^θΛ^N$ of the $N$-particle directed animals ($N\gg 1$) on a simple cubic lattice in a three-dimensional space. Establishing the bijection between the particular configuration of the lattice animal and a class of equivalences of words in the 2D projective locally-free semigroup, we find we find $\ln Λ= \lim_{N\to\infty} \ln Z_N / N$ with $Λ= 2(\sqrt{2}+1) \approx 4.8284$.