Hasil untuk "Animal culture"

Fredrik Jansson

Culture is not just traits but a dynamic system of interdependent beliefs, practices and artefacts embedded in cognitive, social and material structures. Culture evolves as these entities interact, generating path dependence, attractor states and tension, with long-term stability punctuated by rapid systemic transformations. Cultural learning and creativity is modelled as coherence-seeking information processing: individuals filter, transform and recombine input in light of prior acquisitions and dissonance reduction, thereby creating increasingly structured worldviews. Higher-order traits such as goals, skills, norms and cognitive gadgets act as emergent metafilters that regulate subsequent selection by defining what counts as coherent. Together, these filtering processes self-organise into epistemic niches, echo chambers, polarised groups and institutions that channel information flows and constrain future evolution. In this view, LLMs and recommender algorithms are products of cultural embeddings that now act back on cultural systems by automated filtering and recombination of information, reshaping future dynamics of cultural systems.

Suchae Jeong, Inseong Choi, Youngsik Yun et al.

Text-to-Image models, including Stable Diffusion, have significantly improved in generating images that are highly semantically aligned with the given prompts. However, existing models may fail to produce appropriate images for the cultural concepts or objects that are not well known or underrepresented in western cultures, such as `hangari' (Korean utensil). In this paper, we propose a novel approach, Culturally-Aware Text-to-Image Generation with Iterative Prompt Refinement (Culture-TRIP), which refines the prompt in order to improve the alignment of the image with such culture nouns in text-to-image models. Our approach (1) retrieves cultural contexts and visual details related to the culture nouns in the prompt and (2) iteratively refines and evaluates the prompt based on a set of cultural criteria and large language models. The refinement process utilizes the information retrieved from Wikipedia and the Web. Our user survey, conducted with 66 participants from eight different countries demonstrates that our proposed approach enhances the alignment between the images and the prompts. In particular, C-TRIP demonstrates improved alignment between the generated images and underrepresented culture nouns. Resource can be found at https://shane3606.github.io/Culture-TRIP.

Tianyu Xiong, Dayi Tan, Wei Tian

Animal pose estimation is a fundamental task in computer vision, with growing importance in ecological monitoring, behavioral analysis, and intelligent livestock management. Compared to human pose estimation, animal pose estimation is more challenging due to high interspecies morphological diversity, complex body structures, and limited annotated data. In this work, we introduce DiffPose-Animal, a novel diffusion-based framework for top-down animal pose estimation. Unlike traditional heatmap regression methods, DiffPose-Animal reformulates pose estimation as a denoising process under the generative framework of diffusion models. To enhance semantic guidance during keypoint generation, we leverage large language models (LLMs) to extract both global anatomical priors and local keypoint-wise semantics based on species-specific prompts. These textual priors are encoded and fused with image features via cross-attention modules to provide biologically meaningful constraints throughout the denoising process. Additionally, a diffusion-based keypoint decoder is designed to progressively refine pose predictions, improving robustness to occlusion and annotation sparsity. Extensive experiments on public animal pose datasets demonstrate the effectiveness and generalization capability of our method, especially under challenging scenarios with diverse species, cluttered backgrounds, and incomplete keypoints.

Yubo Chen, Di Zhao, Yun Sing Koh et al.

Camera-based animal re-identification (Animal Re-ID) can support wildlife monitoring and precision livestock management in large outdoor environments with limited wireless connectivity. In these settings, inference must run directly on collar tags or low-power edge nodes built around microcontrollers (MCUs), yet most Animal Re-ID models are designed for workstations or servers and are too large for devices with small memory and low-resolution inputs. We propose an on-device framework. First, we characterise the gap between state-of-the-art Animal Re-ID models and MCU-class hardware, showing that straightforward knowledge distillation from large teachers offers limited benefit once memory and input resolution are constrained. Second, guided by this analysis, we design a high-accuracy Animal Re-ID architecture by systematically scaling a CNN-based MobileNetV2 backbone for low-resolution inputs. Third, we evaluate the framework with a real-world dataset and introduce a data-efficient fine-tuning strategy to enable fast adaptation with just three images per animal identity at a new site. Across six public Animal Re-ID datasets, our compact model achieves competitive retrieval accuracy while reducing model size by over two orders of magnitude. On a self-collected cattle dataset, the deployed model performs fully on-device inference with only a small accuracy drop and unchanged Top-1 accuracy relative to its cluster version. We demonstrate that practical, adaptable Animal Re-ID is achievable on MCU-class devices, paving the way for scalable deployment in real field environments.

Chen Cecilia Liu, Anna Korhonen, Iryna Gurevych

Adapting large language models (LLMs) to diverse cultural values is a challenging task, as existing LLMs often reflect the values of specific groups by default, and potentially causing harm to others. In this paper, we present CLCA, a novel framework for enhancing LLM alignment with cultural values based on cultural learning. The framework leverages simulated social interactions to generate conversations in which LLMs engage in role-playing within culturally adapted social scenarios, capturing implicit cultural norms for model fine-tuning. CLCA improves cultural value alignment across various model architectures measured using World Value Survey data, demonstrating the effectiveness of our proposed approach. Our results provide early evidence that understanding intent and social interactions can enhance cultural value adaptation in LLMs, highlighting the promise of training approaches based on cultural learning.

Evan S. Ross, Lance C. Visser, Lalida Tantisuwat et al.

ABSTRACT Background The effects of general anesthesia (GA) on less flow‐dependent (velocity ratio, velocity time integral [VTI] ratio and indexed pulmonary valve area [iPVA]) and flow‐dependent (mean [PVmeanPG] and maximum pressure gradient [PVmaxPG]) indices of severity of pulmonary stenosis (PS) are unclear. Objectives Determine the effects of GA on indices of severity of PS in dogs undergoing an interventional procedure (IP). Determine the reproducibility of indices of severity of PS. Animals Thirty‐nine dogs with PS. Methods Prospective cross‐sectional study. Five repeated echocardiograms were performed over 3 days. Day 1: two echocardiograms were performed by 2 different operators. Day 2: echocardiograms were performed before and after GA but before IP. Day 3: an echocardiogram was performed after the IP. Results After GA, median (IQR) cardiac index (2.1 [1.6–2.6] L/min/m2), PVmeanPG (45.0 [26.0–55.2] mmHg), PVmaxPG (76.6 [46.6–100.3] mmHg) were decreased (p ≤0.001) compared to before GA (2.8 [2.2–3.0] L/min/m2, 55.9 [47.6–73.1] mmHg, 96.1 [81.6–127.0] mmHg, respectively). There were no differences (p ≥0.35) in velocity ratio, VTI ratio, or iPVA after GA. Intra‐operator and inter‐operator coefficients of variation (95% CI) were highest for iPVA (13.8% [10.4–18.4] and 13.5% [11.0–18.4], respectively) and lowest for velocity ratio (9.2% [7.7–12.3] and 9.3% [7.7–12.4], respectively). Conclusions and Clinical Importance PVmeanPG and PVmaxPG might be misleading in states of reduced flow. An integrative assessment of severity of PS that includes less flow‐dependent indices is recommended. Reproducibility of indices of severity of PS should be considered when re‐evaluating dogs with PS.

C.L. van Zyl, H.K. Eriksson, E.A.M. Bokkers et al.

ABSTRACT: In cow-calf contact (CCC) systems breaking the maternal bond may induce stress for the cow, thereby affecting feed intake, milk yield, milk flow rate, and milk electrical conductivity. This study aimed to determine the consequences of weaning and separation strategies in CCC systems for feed intake and milking characteristics of the cow. In 2 experiments, Swedish Holstein and Swedish Red cows either had (experiment 1) whole-day CCC (CCC1, n = 12) for 8.5 ± 1.2 wk (mean ± SD) followed by 12 h of daytime CCC for 8 wk, before abrupt weaning and separation at 16.4 ± 1.2 wk, or (experiment 2) whole-day CCC for 16 ± 1.0 wk; thereafter half of the calves were weaned via nose flaps for 2 wk (NF, n = 10) before physical separation and half via nose flaps for 1 wk and fence-line contact for 1 wk (NFFL, n = 9). Cows were compared with conventionally managed cows (CONV1 or CONV2 in experiment 1 or 2) separated from their calves within 12 h postpartum. In experiment 1, the study period included the week before and after the system switch from whole-day to daytime CCC, and the week before and after separation. In experiment 2, the study period included the week before the start of weaning, during weaning, and 1 week after separation. All cows were milked in the same automatic milking unit. In experiment 1, feed intake of CCC1 cows at separation tended to be lower than CONV1 cows. In experiment 2, roughage intake of NF, NFFL, and CONV2 cows did not differ, but the concentrate intake of NF cows was lower than that of CONV2 cows. In experiment 1, the system switch did not affect milking characteristics. However, after separation, machine milk yield and milk electrical conductivity of CCC1 cows increased, remaining lower than CONV1 cows. In experiment 2, machine milk yield of NF and NFFL cows increased when calves were fitted with nose flaps, but remained lower than CONV2 cows. In the week after separation, milk yield of NFFL cows was similar to that of CONV2 cows, and the NF cows remained lower. In the week before weaning, milk flow rates of NF cows were lower than those of CONV2 cows, and the NFFL cows did not differ. Before weaning, milk electrical conductivity of NF and NFFL cows was lower than that of CONV2 cows, but not thereafter. In conclusion, machine milk yield of CCC cows remained lower either until the week of separation, for NFFL cows, or until 3 or 11 wk after weaning and separation for CCC1 and NF cows of experiments 1 and 2, respectively. Cow-calf contact reduced milk electrical conductivity, and milk and peak milk flow rates increased the week after separation of cow and calf. Not for experiment 2, but for experiment 1, cow roughage and concentrate intake decreased at separation and recovered within a week, indicating that abrupt separation exerted a greater impact on the cow than separation after nose flap weaning or fence-line contact. Future studies should compare both weaning strategies within the same experimental setup, also focusing on the consequences for calves.

Alicia Manzanares-Pedrosa, Joanna Szumilas, Teresa Ayats et al.

Thermophilic Campylobacter spp. are the main cause of gastrointestinal illness in humans through contaminated food. Poultry and poultry products are the main sources of Campylobacter infection. Epidemiological data on Campylobacter prevalence and load in broiler livers remain limited and its presence in this offal may be associated with the caecal load. Hence, this study aimed to determine the prevalence and levels of Campylobacter in chicken livers, both from the surface and inner tissue, compared with that of caeca, by sampling 56 flocks from two slaughterhouses in Spain. Three carcasses per flock were randomly collected during evisceration (n = 168 livers and caecal contents). Overall Campylobacter prevalence was 57.1 % in caecal samples, 77.9 % in surface liver samples and 35.7 % in the inner tissue liver. C. jejuni was the most common species in all sample types and coinfections with C. coli were more prevalent in livers than in the caeca samples. However, there was no relationship between Campylobacter species (C. jejuni, C. coli) and sample type (P > 0.05). The data highlights the role of chicken offal as a potential source of human campylobacteriosis, particularly because of the high Campylobacter load (>103 CFU/liver) in a high proportion of the surface liver samples (40.1 %). However, this high load was only detected in 6.6 % of the inner tissue livers. Restriction fragment length polymorphism (RFLP) analysis revealed a high genetic diversity with 107 different profiles among 473 genotyped Campylobacter isolates. Translocation of Campylobacter strains was demonstrated, with the same RFLP profile identified in isolates from the caeca and the inner liver tissue of the same carcass (14.9 %). Cross-contamination was also revealed, since the same RFLP profile was identified in isolates from the caeca and the surface of the liver from the same carcass (11.9 %). Targeted measures on broiler farms and slaughterhouses to reduce Campylobacter prevalence and cross-contamination in chicken offal will help to reduce the risk of campylobacteriosis for consumers.

Shaoyang Xu, Yongqi Leng, Linhao Yu et al.

As large language models (LLMs) become increasingly accessible in many countries, it is essential to align them to serve pluralistic human values across cultures. However, pluralistic culture alignment in LLMs remain an open problem. In this paper, we propose CultureSPA, a Self-Pluralising Culture Alignment framework that allows LLMs to simultaneously align to pluralistic cultures. The framework first generates questions on various culture topics, then yields LLM outputs in response to these generated questions under both culture-aware and culture-unaware settings. By comparing culture-aware/unaware outputs, we are able to detect and collect culture-related instances. These instances are employed to fine-tune LLMs to serve pluralistic cultures in either a culture-joint or culture-specific way. Extensive experiments demonstrate that CultureSPA significantly improves the alignment of LLMs to diverse cultures without compromising general abilities. And further improvements can be achieved if CultureSPA is combined with advanced prompt engineering techniques. Comparisons between culture-joint and culture-specific tuning strategies, along with variations in data quality and quantity, illustrate the robustness of our method. We also explore the mechanisms underlying CultureSPA and the relations between different cultures it reflects.

Risa Shinoda, Kaede Shiohara

Animal habitat surveys play a critical role in preserving the biodiversity of the land. One of the effective ways to gain insights into animal habitats involves identifying animal footprints, which offers valuable information about species distribution, abundance, and behavior. However, due to the scarcity of animal footprint images, there are no well-maintained public datasets, preventing recent advanced techniques in computer vision from being applied to animal tracking. In this paper, we introduce OpenAnimalTracks dataset, the first publicly available labeled dataset designed to facilitate the automated classification and detection of animal footprints. It contains various footprints from 18 wild animal species. Moreover, we build benchmarks for species classification and detection and show the potential of automated footprint identification with representative classifiers and detection models. We find SwinTransformer achieves a promising classification result, reaching 69.41% in terms of the averaged accuracy. Faster-RCNN achieves mAP of 0.295. We hope our dataset paves the way for automated animal tracking techniques, enhancing our ability to protect and manage biodiversity. Our dataset and code are available at https://github.com/dahlian00/OpenAnimalTracks.

Jingshu Li, Aaditya Patwari, Yi-Chieh Lee

Rises in the number of animal abuse cases are reported around the world. While chatbots have been effective in influencing their users' perceptions and behaviors, little if any research has hitherto explored the design of chatbots that embody animal identities for the purpose of eliciting empathy toward animals. We therefore conducted a mixed-methods experiment to investigate how specific design cues in such chatbots can shape their users' perceptions of both the chatbots' identities and the type of animal they represent. Our findings indicate that such chatbots can significantly increase empathy, improve attitudes, and promote prosocial behavioral intentions toward animals, particularly when they incorporate emotional verbal expressions and authentic details of such animals' lives. These results expand our understanding of chatbots with non-human identities and highlight their potential for use in conservation initiatives, suggesting a promising avenue whereby technology could foster a more informed and empathetic society.

Martin Tancer

A collection of unit cubes with integer coordinates in $\mathbb R^3$ is an animal if its union is homeomorphic to the 3-ball. Pach's animal problem asks whether any animal can be transformed to a single cube by adding or removing cubes one by one in such a way that any intermediate step is an animal as well. Here we provide an example of an animal that cannot be transformed to a single cube this way within its bounding box.

Gabriela de Assis dos Santos, João Manoel Magalhães Almeida Bezerra, Ana Beatriz Santana Silva et al.

O megaesôfago é uma enfermidade ocasionada pela hipomotilidade e dilatação esofágica parcial ou total, podendo ser congênito ou adquirido e subdivididos em primário, secundário, ou idiopático. Na forma congênita a patogenia não está totalmente esclarecida, mas, a sintomatologia inicia-se após o desmame e o filhote da ninhada apresenta subdesenvolvimento. O principal sinal clínico desta doença é a regurgitação após a ingestão de alimento e de água. O diagnóstico definitivo baseia-se no histórico, exame clínico e exames complementares de imagem. O tratamento é conservador, sendo necessário mudar o manejo alimentar e tratar as complicações que podem ser ocasionadas. Além disso, os cães mais acometidos são os médio a grande porte. Sendo assim, o objetivo deste trabalho é relatar um caso de um megaesôfago congênito em cão da raça Pinscher, filhote, com histórico de regurgitação desde o desmame, apetite voraz e retardo no desenvolvimento. Após o diagnóstico através de radiografia simples e contrastada foi iniciado o manejo alimentar e o paciente demonstrou melhora clínica.

Matthew Elliott, Benjamin Golub, Matthieu V. Leduc

Complex organizations accomplish tasks through many steps of collaboration among workers. Corporate culture supports collaborations by establishing norms and reducing misunderstandings. Because a strong corporate culture relies on costly, voluntary investments by many workers, we model it as an organizational public good, subject to standard free-riding problems, which become severe in large organizations. Our main finding is that voluntary contributions to culture can nevertheless be sustained, because an organization's equilibrium productivity is endogenously highly sensitive to individual contributions. However, the completion of complex tasks is then necessarily fragile to small shocks that damage the organization's culture.

Zhangsihao Yang, Mingyuan Zhou, Mengyi Shan et al.

Our paper aims to generate diverse and realistic animal motion sequences from textual descriptions, without a large-scale animal text-motion dataset. While the task of text-driven human motion synthesis is already extensively studied and benchmarked, it remains challenging to transfer this success to other skeleton structures with limited data. In this work, we design a model architecture that imitates Generative Pretraining Transformer (GPT), utilizing prior knowledge learned from human data to the animal domain. We jointly train motion autoencoders for both animal and human motions and at the same time optimize through the similarity scores among human motion encoding, animal motion encoding, and text CLIP embedding. Presenting the first solution to this problem, we are able to generate animal motions with high diversity and fidelity, quantitatively and qualitatively outperforming the results of training human motion generation baselines on animal data. Additionally, we introduce AnimalML3D, the first text-animal motion dataset with 1240 animation sequences spanning 36 different animal identities. We hope this dataset would mediate the data scarcity problem in text-driven animal motion generation, providing a new playground for the research community.

Yan Tao, Olga Viberg, Ryan S. Baker et al.

Culture fundamentally shapes people's reasoning, behavior, and communication. As people increasingly use generative artificial intelligence (AI) to expedite and automate personal and professional tasks, cultural values embedded in AI models may bias people's authentic expression and contribute to the dominance of certain cultures. We conduct a disaggregated evaluation of cultural bias for five widely used large language models (OpenAI's GPT-4o/4-turbo/4/3.5-turbo/3) by comparing the models' responses to nationally representative survey data. All models exhibit cultural values resembling English-speaking and Protestant European countries. We test cultural prompting as a control strategy to increase cultural alignment for each country/territory. For recent models (GPT-4, 4-turbo, 4o), this improves the cultural alignment of the models' output for 71-81% of countries and territories. We suggest using cultural prompting and ongoing evaluation to reduce cultural bias in the output of generative AI.

Mor Ventura, Eyal Ben-David, Anna Korhonen et al.

Text-To-Image (TTI) models, such as DALL-E and StableDiffusion, have demonstrated remarkable prompt-based image generation capabilities. Multilingual encoders may have a substantial impact on the cultural agency of these models, as language is a conduit of culture. In this study, we explore the cultural perception embedded in TTI models by characterizing culture across three hierarchical tiers: cultural dimensions, cultural domains, and cultural concepts. Based on this ontology, we derive prompt templates to unlock the cultural knowledge in TTI models, and propose a comprehensive suite of evaluation techniques, including intrinsic evaluations using the CLIP space, extrinsic evaluations with a Visual-Question-Answer (VQA) model and human assessments, to evaluate the cultural content of TTI-generated images. To bolster our research, we introduce the CulText2I dataset, derived from six diverse TTI models and spanning ten languages. Our experiments provide insights regarding Do, What, Which and How research questions about the nature of cultural encoding in TTI models, paving the way for cross-cultural applications of these models.

María Angelica Tamame, Viviana Cenizo, José Luis Pall

Las colecciones biológicas son repositorios sistematizados de algún tipo de material biológico que proveen información de su identidad y procedencia en un tiempo determinado. Son documentos de los recursos naturales y constituyen el medio para el conocimiento de la biodiversidad y su conservación. La mayoría de estos repositorios se encuentran en museos de ciencia, universidades, centros de investigación y colecciones privadas. En particular, las palinotecas son colecciones de referencia de esporas, granos de polen y otros palinomorfos, que incluyen preparados microscópicos y son de gran importancia para la identificación de los taxones productores. El objetivo de este trabajo fue organizar, actualizar y clasificar el material de la colección palinológica de la Facultad de Agronomía de la Universidad Nacional de La Pampa. Se ordenó y actualizó tanto la base de datos como el material físico que incluía preparados microscópicos, tubos con residuos polínicos y sobres con restos florales de material herborizado. Estas actividades se realizaron para la palinoteca de dos colecciones, la de material acetolizado (FALPPA) y la de material fresco (FALPPF). Se registraron 1382 preparados correspondientes a 495 taxones y 79 familias botánicas para la colección de material acetolizado. De acuerdo con el estatus de los taxones, el 75,2 % de los preparados palinológicos correspondió a especies nativas de Sudamérica y el 22,6 % a especies exóticas. En cuanto a la procedencia, el 92,3 % de los preparados provino de La Pampa y el 95,5 % de los departamentos de la provincia contaron con ejemplares en la palinoteca, principalmente Lihuel Calel con 712 preparados. Según el origen fitogeográfico, el 72 % correspondió a taxones del Monte o arbustal, el 23 % al Espinal o bosque de caldén y el resto a la estepa graminosa o Provincia Pampeana. Para la colección de material fresco se registraron 68 familias correspondientes a 161 géneros, el 97,8 % de los mismos provinieron del Departamento Capital. La actualización de la palinoteca de referencia permitió organizar la información e identificar aspectos de importancia científica, potenciando el valor de una de las colecciones de la Universidad Nacional de La Pampa. 

Kim Youwang, Kim Ji-Yeon, Kyungdon Joo et al.

We propose an end-to-end unified 3D mesh recovery of humans and quadruped animals trained in a weakly-supervised way. Unlike recent work focusing on a single target class only, we aim to recover 3D mesh of broader classes with a single multi-task model. However, there exists no dataset that can directly enable multi-task learning due to the absence of both human and animal annotations for a single object, e.g., a human image does not have animal pose annotations; thus, we have to devise a new way to exploit heterogeneous datasets. To make the unstable disjoint multi-task learning jointly trainable, we propose to exploit the morphological similarity between humans and animals, motivated by animal exercise where humans imitate animal poses. We realize the morphological similarity by semantic correspondences, called sub-keypoint, which enables joint training of human and animal mesh regression branches. Besides, we propose class-sensitive regularization methods to avoid a mean-shape bias and to improve the distinctiveness across multi-classes. Our method performs favorably against recent uni-modal models on various human and animal datasets while being far more compact.

Hang Yu, Yufei Xu, Jing Zhang et al.

Accurate animal pose estimation is an essential step towards understanding animal behavior, and can potentially benefit many downstream applications, such as wildlife conservation. Previous works only focus on specific animals while ignoring the diversity of animal species, limiting the generalization ability. In this paper, we propose AP-10K, the first large-scale benchmark for mammal animal pose estimation, to facilitate the research in animal pose estimation. AP-10K consists of 10,015 images collected and filtered from 23 animal families and 54 species following the taxonomic rank and high-quality keypoint annotations labeled and checked manually. Based on AP-10K, we benchmark representative pose estimation models on the following three tracks: (1) supervised learning for animal pose estimation, (2) cross-domain transfer learning from human pose estimation to animal pose estimation, and (3) intra- and inter-family domain generalization for unseen animals. The experimental results provide sound empirical evidence on the superiority of learning from diverse animals species in terms of both accuracy and generalization ability. It opens new directions for facilitating future research in animal pose estimation. AP-10k is publicly available at https://github.com/AlexTheBad/AP10K.