Nikhil Reddy Varimalla, Yunfei Xu, Arkadiy Saakyan
et al.
As Video Large Language Models (VideoLLMs) are deployed globally, they require understanding of and grounding in the relevant cultural background. To properly assess these models' cultural awareness, adequate benchmarks are needed. We introduce VideoNorms, a benchmark of over 1000 (video clip, norm) pairs from US and Chinese cultures annotated with socio-cultural norms grounded in speech act theory, norm adherence and violations labels, and verbal and non-verbal evidence. To build VideoNorms, we use a human-AI collaboration framework, where a teacher model using theoretically-grounded prompting provides candidate annotations and a set of trained human experts validate and correct the annotations. We benchmark a variety of open-weight VideoLLMs on the new dataset which highlight several common trends: 1) models performs worse on norm violation than adherence; 2) models perform worse w.r.t Chinese culture compared to the US culture; 3) models have more difficulty in providing non-verbal evidence compared to verbal for the norm adhere/violation label and struggle to identify the exact norm corresponding to a speech-act; and 4) unlike humans, models perform worse in formal, non-humorous contexts. Our findings emphasize the need for culturally-grounded video language model training - a gap our benchmark and framework begin to address.
Mohamed Gamil, Abdelrahman Elsayed, Abdelrahman Lila
et al.
Despite recent advances in AI, multimodal culturally diverse datasets are still limited, particularly for regions in the Middle East and Africa. In this paper, we introduce EgMM-Corpus, a multimodal dataset dedicated to Egyptian culture. By designing and running a new data collection pipeline, we collected over 3,000 images, covering 313 concepts across landmarks, food, and folklore. Each entry in the dataset is manually validated for cultural authenticity and multimodal coherence. EgMM-Corpus aims to provide a reliable resource for evaluating and training vision-language models in an Egyptian cultural context. We further evaluate the zero-shot performance of Contrastive Language-Image Pre-training CLIP on EgMM-Corpus, on which it achieves 21.2% Top-1 accuracy and 36.4% Top-5 accuracy in classification. These results underscore the existing cultural bias in large-scale vision-language models and demonstrate the importance of EgMM-Corpus as a benchmark for developing culturally aware models.
Alistair Plum, Anne-Marie Lutgen, Christoph Purschke
et al.
Large language models have become the latest trend in natural language processing, heavily featuring in the digital tools we use every day. However, their replies often reflect a narrow cultural viewpoint that overlooks the diversity of global users. This missing capability could be referred to as cultural reasoning, which we define here as the capacity of a model to recognise culture-specific knowledge values and social norms, and to adjust its output so that it aligns with the expectations of individual users. Because culture shapes interpretation, emotional resonance, and acceptable behaviour, cultural reasoning is essential for identity-aware AI. When this capacity is limited or absent, models can sustain stereotypes, ignore minority perspectives, erode trust, and perpetuate hate. Recent empirical studies strongly suggest that current models default to Western norms when judging moral dilemmas, interpreting idioms, or offering advice, and that fine-tuning on survey data only partly reduces this tendency. The present evaluation methods mainly report static accuracy scores and thus fail to capture adaptive reasoning in context. Although broader datasets can help, they cannot alone ensure genuine cultural competence. Therefore, we argue that cultural reasoning must be treated as a foundational capability alongside factual accuracy and linguistic coherence. By clarifying the concept and outlining initial directions for its assessment, a foundation is laid for future systems to be able to respond with greater sensitivity to the complex fabric of human culture.
Culture shapes the objects people use and for what purposes, yet mainstream Vision-Language (VL) datasets frequently exhibit cultural biases, disproportionately favoring higher-income, Western contexts. This imbalance reduces model generalizability and perpetuates performance disparities, especially impacting lower-income and non-Western communities. To address these disparities, we propose a novel function-centric framework that categorizes objects by the functions they fulfill, across diverse cultural and economic contexts. We implement this framework by creating the Culture Affordance Atlas, a re-annotated and culturally grounded restructuring of the Dollar Street dataset spanning 46 functions and 288 objects publicly available at https://lit.eecs.umich.edu/CultureAffordance-Atlas/index.html. Through extensive empirical analyses using the CLIP model, we demonstrate that function-centric labels substantially reduce socioeconomic performance gaps between high- and low-income groups by a median of 6 pp (statistically significant), improving model effectiveness for lower-income contexts. Furthermore, our analyses reveals numerous culturally essential objects that are frequently overlooked in prominent VL datasets. Our contributions offer a scalable pathway toward building inclusive VL datasets and equitable AI systems.
Simon Ravé, Jean-Christophe Lombardo, Pejman Rasti
et al.
We present a zero-shot segmentation approach for agricultural imagery that leverages Plantnet, a large-scale plant classification model, in conjunction with its DinoV2 backbone and the Segment Anything Model (SAM). Rather than collecting and annotating new datasets, our method exploits Plantnet's specialized plant representations to identify plant regions and produce coarse segmentation masks. These masks are then refined by SAM to yield detailed segmentations. We evaluate on four publicly available datasets of various complexity in terms of contrast including some where the limited size of the training data and complex field conditions often hinder purely supervised methods. Our results show consistent performance gains when using Plantnet-fine-tuned DinoV2 over the base DinoV2 model, as measured by the Jaccard Index (IoU). These findings highlight the potential of combining foundation models with specialized plant-centric models to alleviate the annotation bottleneck and enable effective segmentation in diverse agricultural scenarios.
J. I. Ruiz-Martinez, A. Mendez-Vazquez, E. Rodriguez-Tello
In precision agriculture, one of the most important tasks when exploring crop production is identifying individual plant components. There are several attempts to accomplish this task by the use of traditional 2D imaging, 3D reconstructions, and Convolutional Neural Networks (CNN). However, they have several drawbacks when processing 3D data and identifying individual plant components. Therefore, in this work, we propose a novel Deep Learning architecture to detect components of individual plants on Light Detection and Ranging (LiDAR) 3D Point Cloud (PC) data sets. This architecture is based on the concept of Graph Neural Networks (GNN), and feature enhancing with Principal Component Analysis (PCA). For this, each point is taken as a vertex and by the use of a K-Nearest Neighbors (KNN) layer, the edges are established, thus representing the 3D PC data set. Subsequently, Edge-Conv layers are used to further increase the features of each point. Finally, Graph Attention Networks (GAT) are applied to classify visible phenotypic components of the plant, such as the leaf, stem, and soil. This study demonstrates that our graph-based deep learning approach enhances segmentation accuracy for identifying individual plant components, achieving percentages above 80% in the IoU average, thus outperforming other existing models based on point clouds.
Silicon has striking similarity with carbon and is found in plant cells. However, there is no specific role that has been assigned to silicon in the life cycle of plants. The amount of silicon in plant cells is species specific and can reach levels comparable to macronutrients. Silicon is the central element for artificial intelligence, nanotechnology and digital revolution thus can act as an informational molecule like nucleic acids while the diverse bonding potential of silicon with different chemical species is analogous to carbon and thus can serve as a structural candidate such as proteins. The discovery of large amounts of silicon on Mars and the moon along with the recent developments of enzyme that can incorporate silicon into organic molecules has propelled the theory of creating silicon-based life. More recently, bacterial cytochrome has been modified through directed evolution such that it could cleave silicon-carbon bonds in organo-silicon compounds thus consolidating on the idea of utilizing silicon in biomolecules. In this article the potential of silicon-based life forms has been hypothesized along with the reasoning that autotrophic virus-like particles can be a lucrative candidate to investigate such potential. Such investigations in the field of synthetic biology and astrobiology will have corollary benefit on Earth in the areas of medicine, sustainable agriculture and environmental sustainability. Bibliometric analysis indicates an increasing interest in synthetic biology. Germany leads in research related to plant synthetic biology, while Biotechnology and Biological Sciences Research Council (BBSRC) at UK has highest financial commitments and Chinese Academy of Sciences generates the highest number of publications in the field.
Accurate point cloud segmentation for plant organs is crucial for 3D plant phenotyping. Existing solutions are designed problem-specific with a focus on certain plant species or specified sensor-modalities for data acquisition. Furthermore, it is common to use extensive pre-processing and down-sample the plant point clouds to meet hardware or neural network input size requirements. We propose a simple, yet effective algorithm KDSS for sub-sampling of biological point clouds that is agnostic to sensor data and plant species. The main benefit of this approach is that we do not need to down-sample our input data and thus, enable segmentation of the full-resolution point cloud. Combining KD-SS with current state-of-the-art segmentation models shows satisfying results evaluated on different modalities such as photogrammetry, laser triangulation and LiDAR for various plant species. We propose KD-SS as lightweight resolution-retaining alternative to intensive pre-processing and down-sampling methods for plant organ segmentation regardless of used species and sensor modality.
Olea europaea L., olive tree, has a very important role in the economy of the Mediterranean region, where 93 % of the world's olive oil is produced. This species is well adapted to the environmental conditions of this area, but the increase in the frequency of extreme climatic events, due to climate change, is affecting the yield and quality of olive products. The use of eco-friendly solutions, like plant-beneficial microorganisms, can be a sustainable agronomic tool to improve plant tolerance to stress and boost agricultural production. We aim to unravel the effects of the pre-treatment of olive plants with the bacterium Pseudomonas reactans Ph3R3 on drought tolerance. Young potted olive plants were treated with a solution of P. reactans (soil inoculation) or with distilled water, and then exposed to two watering conditions (well-watered or water deficit). Plant water status, photosynthesis, pigments, carbohydrates, oxidative stress biomarkers, and total antioxidant capacity were evaluated 61 and 191 days after the beginning of the watering treatments. The pre-treatment with P. reactans improved leaf dry biomass production, and soil C and N availability. Moreover, under drought conditions, P. reactans increased leaf water availability, N levels, and the intercellular CO2, leading to improved net CO2 assimilation rate and carbohydrates production. Also, P. reactans activated stress protective strategies (total antioxidant capacity) that helped to control oxidative stress. These data demonstrated that the benefits triggered by P. reactans pre-treatment promoted olive performance and tolerance to drought and could be a promising strategy to improve olive culture sustainability.
Vanesa Raya, Margarita Parra, María del Carmen Cid
et al.
Defoliation (leaf removal or pruning) is a common practice in tomato production that makes crops more manageable, prevents conditions conducive to fungal attack and increases the exposure of the fruit to light, especially in winter conditions. The intensity and frequency of leaf removal on commercial farms often vary according to workforce availability criteria, which makes it difficult to determine their effect on tomato crop yields. It would be reasonable to think that a reduction in leaf area influences radiation interception and, therefore, the production of assimilates and biomass. However, in intensive production systems with a high leaf area index (LAI), leaf pruning can increase radiation interception, either by reducing competition between productive and vegetative organs or by increasing radiation use efficiency. This study was therefore designed to assess the effect of different intensities and frequencies of basal leaf removal on dry matter production and partitioning between the different organs of the plant, and thus on tomato crop productivity. A series of trials were conducted over three consecutive seasons, with a trial conducted per season: (a) Trial 1: leaf removal control—LRC (with leaves removed from the base to two leaves below the truss close to harvest, T0) was compared with LR1 (leaf removal from the base to two leaves below the truss above T0, i.e., T1) and LR2 (two trusses above T0 (T2)); (b) Trial 2: LRC compared with LR2 and LR4 (four trusses above T0 (T4)), carried out at two frequencies; and (c) Trial 3: LRC compared with an intense leaf removal treatment (LRI) whereby between 10 and 12 leaves were left on each stem. LAI saturation values under our conditions were found to be around 2.0. No significant differences in yield were found between the control and treatments LR1, LR2 and LR4, with a reduction in the number of leaves of up to 35% and LAI values during harvest above 2.0. The intense leaf removal treatment (LRI), which reduced the number of leaves by 47% and the LAI value from 2.8 to 1.5 compared to the control, resulted in a 15% reduction in dry biomass and a 17% decrease in fruit yield.
Gary W. Knox, Edward F. Gilman, Teagan Young
et al.
Pruning is one of the most controversial aspects of maintaining crapemyrtle. Traditionally, many crapemyrtles were routinely topped, leaving large branch and stem stubs. This practice has been called "crape murder" because most people dislike the winter appearance and many professionals believe the practice impacts crapemyrtle health and structural integrity. UF/IFAS research has clarified the effects of various crapemyrtle pruning practices that resulted in the recommendations in this publication.
Olena Burda-Lassen, Aman Chadha, Shashank Goswami
et al.
An image is often considered worth a thousand words, and certain images can tell rich and insightful stories. Can these stories be told via image captioning? Images from folklore genres, such as mythology, folk dance, cultural signs, and symbols, are vital to every culture. Our research compares the performance of four popular vision-language models (GPT-4V, Gemini Pro Vision, LLaVA, and OpenFlamingo) in identifying culturally specific information in such images and creating accurate and culturally sensitive image captions. We also propose a new evaluation metric, the Cultural Awareness Score (CAS), which measures the degree of cultural awareness in image captions. We provide a dataset MOSAIC-1.5k labeled with ground truth for images containing cultural background and context and a labeled dataset with assigned Cultural Awareness Scores that can be used with unseen data. Creating culturally appropriate image captions is valuable for scientific research and can be beneficial for many practical applications. We envision our work will promote a deeper integration of cultural sensitivity in AI applications worldwide. By making the dataset and Cultural Awareness Score available to the public, we aim to facilitate further research in this area, encouraging the development of more culturally aware AI systems that respect and celebrate global diversity.
In this paper we delve into the historical evolution of data as a fundamental element in communication and knowledge transmission. The paper traces the stages of knowledge dissemination from oral traditions to the digital era, highlighting the significance of languages and cultural diversity in this progression. It also explores the impact of digital technologies on memory, communication, and cultural preservation, emphasizing the need for promoting a culture of the digital (rather than a digital culture) in Africa and beyond. Additionally, it discusses the challenges and opportunities presented by data biases in AI development, underscoring the importance of creating diverse datasets for equitable representation. We advocate for investing in data as a crucial raw material for fostering digital literacy, economic development, and, above all, cultural preservation in the digital age.
Eugene Koh, Rohan Shawn Sunil, Hilbert Yuen In Lam
et al.
Life finds a way. For sessile organisms like plants, the need to adapt to changes in the environment is even more poignant. For humanity, the need to develop crops that can grow in diverse environments and feed our growing population is an existential one. The advent of the genomics era enabled the generation of high-throughput data and computational methods that serve as powerful hypothesis-generating tools to understand the genomic and gene functional basis of stress resilience. Today, the proliferation of artificial intelligence (AI) allows scientists to rapidly screen through high-throughput datasets to uncover elusive patterns and correlations, enabling us to create more performant models for prediction and hypothesis generation in plant biology. This review aims to provide an overview of the availability of large-scale data in plant stress research and discuss the application of AI tools on these large-scale datasets in a bid to develop more stress-resilient plants.
The Munake grape is a local variety of grape that is widely distributed in Xinjiang, China. This study aims to clarify the genetic structure of the Munake grape population, characterize genetic differentiation and gene flow among populations, gather germplasm, and establish the core germplasm collection. In total, 144 samples were collected from eight geographic populations. Twenty-two SSR markers were used to characterize the genetic diversity as well as the genetic structure of Munake grape germplasm and to establish the core germplasm collection. At each site, the average number of effective alleles (Ne) was 5.019. Overall, genetic diversity was high in the various geographic populations of Munake grapes. Polymorphic information content (PIC) ranged from 0.501 to 0.908, with an average of 0.728. Estimates of genetic differentiation and gene flow indicated that the Artux population had significant genetic differences from the other populations. Screening results indicated that a sampling proportion of 95% of the sample was required to achieve 100% allelic coverage, or a sampling proportion of 65% for 95% allelic coverage. This analysis was based on conventional genetic diversity indicators, with a core germplasm diversity index of 95% coverage. Characterization of the genetic diversity of germplasm from 144 Munake grapes not only provides valuable resources for future genetic mapping and functional genome research, but also facilitates the utilization of core germplasm and molecular breeding of Munake grapes.
Heat shock transcription factors (Hsfs) play important roles in plant developmental regulations and various stress responses. In present study, 46 Hsf genes in peanut (AhHsf) were identified and analyzed. The 46 AhHsf genes were classed into three groups (A, B, and C) and 14 subgroups (A1-A9, B1-B4, and C1) together with their Arabidopsis homologs according to phylogenetic analyses, and 46 AhHsf genes unequally located on 17 chromosomes. Gene structure and protein motif analysis revealed that members from the same subgroup possessed similar exon/intron and motif organization, further supporting the results of phylogenetic analyses. Gene duplication events were found in peanut Hsf gene family via syntenic analysis, which were important in Hsf gene family expansion in peanut. The expression of AhHsf genes were detected in different tissues using published data, implying that AhHsf genes may differ in function. In addition, several AhHsf genes (AhHsf5, AhHsf11, AhHsf20, AhHsf24, AhHsf30, AhHsf35) were induced by drought and salt stresses. Furthermore, the stress-induced member AhHsf20 was found to be located in nucleus. Notably, overexpression of AhHsf20 was able to enhance salt tolerance. These results from this study may provide valuable information for further functional analysis of peanut Hsf genes.
In order to make the non-heritage culture of Yicheng Flower Drum more relevant to the trend of the digital era and promote its dissemination and inheritance, the design and application of gesture recognition and virtual reality technologies guided by embodied cognition theory in the process of non-heritage culture dissemination is studied. At the same time, it will enhance the interaction between people and NRM culture, stimulate the audience's interest in understanding NRM and spreading NRM, and create awareness of preserving NRM culture. Using embodied cognition as a theoretical guide, expanding the unidirectional communication mode through human-computer interaction close to natural behavior and cooperating with multisensory information reception channels, so as to construct an embodied and immersive interactive atmosphere for the participants and enable them to naturally form the cognition and understanding of the traditional culture in the process of interaction. The dissemination of the non-heritage culture Yicheng Flower Drum can take the theory of embodied cognition as an entry point, and through the virtual and real scenes of Yicheng Flower Drum and the immersive experience, we can empower the interaction design of non-heritage culture dissemination of the virtual and real, and provide a new method for the research of digital design of non-heritage culture.
Ashley R. Smyth, Laura K. Reynolds, Savanna C. Barry
et al.
The purpose of this new 6-page document is to explain the types of ecosystem services provided by different types of living shorelines and how to quantify these values. The target audience for this document is local governments and municipalities that make decisions about developing, conserving, and restoring living shorelines; state management agencies that oversee broader scale habitat management; and finally, homeowners who will be immediately affected by any of these decisions. Written by Ashley R. Smyth, Laura K. Reynolds, Savanna C. Barry, Natalie C. Stephens, Joshua T. Patterson, and Edward V. Camp and published by the UF/IFAS Department of Soil, Water, and Ecosystem Sciences.
https://edis.ifas.ufl.edu/SS707
Drought and salt stress can strongly affect the growth and development of wheat. Wheat adapts to drought and salt stress through osmotic regulation. Betaine aldehyde dehydrogenase (BADH) is a key enzyme in the synthesis of betaine, an osmotic regulator. We cloned a region of the TaBADH-A1 promoter and genomic DNA that included the introns and exons, from four Chinese wheat cultivars. Following the analysis of TaBADH-A1 genomic DNA and promoter sequence polymorphisms of 4 cloned and 15 cultivars from the database, 7 haplotypes of TaBADH-A1 gene were identified. We divided the 7 haplotypes with a 254 bp insertion or deletion (indel) into two main alleles, BADH-A1a and BADH-A1b. Meanwhile, a molecular marker was developed based on the 254 bp indel of the third intron of TaBADH-A1 gene. Expression levels of BADH-A1b were found to be significantly higher than those of BADH-A1a under drought and salt stress conditions. Betaine accumulation was significantly higher in wheat containing BADH-A1b compared to BADH-A1a under drought and salt stress. We also identified that the average relative germination and survival rates of wheat with the BADH-A1b allele were significantly higher than wheat with the BADH-A1a allele. The results reveal that wheat containing BADH-A1b has stronger drought and salt tolerance than wheat with BADH-A1a. Meanwhile, the geographic distribution and frequency of the TaBADH-A1 locus alleles indicate that BADH-A1a has been preferred in Chinese wheat breeding programs, while BADH-A1b, associated with favorable stress tolerance, has been neglected. The results of this study provide evidence for an excellent candidate allele for marker-assisted selection of new wheat cultivars with increased salt tolerance and drought resistance.