Political scientists are increasingly interested in analyzing visual content at scale. However, the existing computational toolbox is still in need of methods and models attuned to the specific challenges and goals of social and political inquiry. In this article, we introduce a visual structural topic model (vSTM) that combines continuous image embeddings with a structural topic model. This has important advantages compared to existing approaches. First, continuous embeddings allow the model to capture the semantic complexity of images, which is relevant to the study of politics. A consequence of using continuous embeddings is that we can directly use off-the-shelf pretrained deep learning models, offering state-of-the-art performance in image analysis tasks while saving time, cost and resources. Second, the structural topic model provides the ability to analyze how topics and covariates are related, while maintaining a nuanced representation of images as a mixture of multiple topics. In our experiments, we show that the vSTM is able to identify topics that are interpretable, coherent and relevant to the study of online political communication.

As computer vision becomes more popular in archaeology, it is imperative to develop best practices that balance the interpretive and analytically critical tendencies of manual survey with the systemization and efficiency promised by machine learning. Using a case study from the northern US Southwest, we demonstrate an iterative image classification approach that reflects the methodological processes that play out in field research and preserves the role of humans as nuanced decision-makers. We develop a model to identify ancestral Pueblo residential sites and use it to survey an area larger than 16,000 km2. The semiautomated survey identified 4,905 likely archaeological features, marking one of the largest remote archaeological surveys in North America. Results not only emphasize the value of integrated computer vision for archaeological survey and site prediction but also demonstrate a research design that capitalizes on the computational value of computer vision while maintaining active engagement by the researcher.

Construction safety inspections typically involve a human inspector identifying safety concerns on-site. With the rise of powerful vision language models (VLMs), researchers are exploring their use for tasks such as detecting safety rule violations from on-site images. However, there is a lack of open datasets to comprehensively evaluate and further fine-tune VLMs in construction safety inspection. Current applications of VLMs use small, supervised datasets, limiting their applicability in tasks they are not directly trained for. In this article, we propose the ConstructionSite 10 k, featuring 10,000 construction site images with annotations for three inter-connected tasks, including image captioning, safety rule violation visual question answering (VQA), and construction element visual grounding. Our subsequent evaluation of current state-of-the-art large pre-trained VLMs shows notable generalization abilities in zero-shot and few-shot settings, while additional training is needed to make them applicable to actual construction sites. This dataset allows researchers to train and evaluate their own VLMs with new architectures and techniques, providing a valuable benchmark for construction safety inspection.

Why do politicians sometimes deliver passionate speeches and sometimes tedious monologues? Even though the delivery is key to understanding political speech, we know little about when and why political actors choose particular delivery styles. Focusing on legislative speech, we expect legislators to deliver more emphatic speeches when their vote is aligned with the preferences of their constituents. To test this proposition, we develop and apply an automated video analysis model to speech recordings from the US House of Representatives. We match the speech emphasis with district preferences on key bills using data from the Cooperative Congressional Election Study. We find that House members who rise in opposition to a bill give more passionate speeches when public preferences are aligned with their vote. The results suggest that political actors are not only mindful of public opinion in what they say but also in how they say it.

Female genital schistosomiasis (FGS) is a chronically disabling gynaecological condition, impacting up to 56 million women and girls, mostly in sub-Saharan Africa. In lieu of a gold standard laboratory test, it is possible to diagnose FGS visually. Visual diagnosis is performed through inspection of the cervix and surrounding tissue to identify signs of Schistosoma egg deposition, associated inflammation and granuloma formation. The change related to egg deposition can be very subtle and heterogeneous and is often seen in the context of other altered cervical morphology. Visual diagnostics for FGS are therefore currently highly subjective and lack specificity, with low consistency of grading between trained expert reviewers. Computer vision, driven by artificial intelligence, is an enticing prospect to overcome these issues due to the potential to accurately detect and classify the subtle changes and patterns that are indiscernible to human graders. Computer vision also offers the opportunity to support resource-constrained regions with few staff trained on visual diagnostics. However, several challenges stand in the way of progressing and successfully implementing computer vision tools for FGS. These challenges are particularly related to the variation in the appearance of the cervix (with or without disease) and FGS lesions, as well as the difficulty with accurately labelling cervical images. Exploring alternative annotation methods and model architectures is likely to improve the performance of FGS computer vision tools. This paper will explore the challenges of FGS computer vision and provide suggestions on how to overcome these barriers to enhance visual diagnostics for FGS.

Research in political science has begun to explore how to use large language and object detection models to analyze text and visual data. However, few studies have explored how to use these tools for data extraction. Instead, researchers interested in extracting text from poorly formatted sources typically rely on optical character recognition and regular expressions or extract each item by hand. This letter describes a workflow process for structured text extraction using free models and software. I discuss the type of data best suited to this method, its usefulness within political science, and the steps required to convert the text into a usable dataset. Finally, I demonstrate the method by extracting agenda items from city council meeting minutes. I find the method can accurately extract subsections of text from a document and requires only a few hand labeled documents to adequately train.

Automatic precision herbicide application offers significant potential for reducing herbicide use in turfgrass weed management. However, developing accurate and reliable neural network models is crucial for achieving optimal precision weed control. The reported neural network models in previous research have been limited by specific geographic regions, weed species, and turfgrass management practices, restricting their broader applicability. The objective of this research was to evaluate the feasibility of deploying a single, robust model for weed classification across a diverse range of weed species, considering variations in species, ecotypes, densities, and growth stages in bermudagrass turfgrass systems across different regions in both China and the United States. Among the models tested, ResNeXt152 emerged as the top performer, demonstrating strong weed detection capabilities across 24 geographic locations and effectively identifying 14 weed species under varied conditions. Notably, the ResNeXt152 model achieved an F1 score and recall exceeding 0.99 across multiple testing scenarios, with a Matthews correlation coefficient (MCC) value surpassing 0.98, indicating its high effectiveness and reliability. These findings suggest that a single neural network model can reliably detect a wide range of weed species in diverse turf regimes, significantly reducing the costs associated with model training and confirming the feasibility of using one model for precision weed control across different turf settings and broad geographic regions.

Weed diversity plays an important role in the functioning of agroecosystems. Moreover, a number of endangered/threatened plant species occur as weeds in arable fields and/or field boundaries. Agricultural intensification has imposed negative consequences on weed diversity in general, and the survival of the endangered/threatened plant species in particular. The objective of this review is to provide a theoretical framework for promoting cropland weed diversity through precision agriculture. A systematic review was conducted based on literature analysis, existing knowledge gaps, and current needs to identify a suitable approach for promoting cropland biodiversity while protecting crop yields. While nonchemical weed management methods and economic threshold–based approaches are touted to improve weed diversity, they are either ineffective or insufficient for this purpose; long-term economic consequences and the risk of weed adaptation are major concerns. A plant functional trait-based approach to promoting weed diversity, one that considers a plant’s ecosystem service potential and competitiveness with the crop, among other factors, has been proposed by researchers. This approach has tremendous potential for weed diversity conservation in commercial production systems, but field implementation has been limited thus far due to our inability to selectively control weeds at the individual-plant level. However, recent advancements in computer vision, machine learning, and site-specific weed management technologies may allow for the accurate elimination of unwanted plants while retaining the important ones. Here, we present a novel framework for the utilization of precision agriculture for the conservation of cropland weed diversity, including the protection of endangered/threatened plant species, while protecting crop yields. This approach is the first of its kind in which the control priority is ranked on an individual-plant basis, by integrating intrinsic weed trait values with field infestation characteristics, while management thresholds are tailored to specific goals and priorities.

Posture-related musculoskeletal issues among office workers are a significant health concern, mainly due to long periods spent in static positions. This research presents a Posture Lab which is a workplace-based solution through an easy-to-use posture monitoring system, allowing employees to assess their posture. The Posture Lab focuses on two key aspects: Normal Head Posture (NHP) versus Forward Head Posture (FHP) measurement and thoracic spine kyphosis. Craniovertebral (CA) and Shoulder Angles (SA) quantify NHP and FHP. The Kyphosis Angle (KA) is for measuring normal thoracic spine and kyphosis. To measure these angles, the system uses computer vision technology with ArUco markers detection via a webcam to analyze head positions. Additionally, wearable accelerometer sensors measure kyphosis by checking the angles of inclination. The framework includes a web-based user interface for registration and specialized desktop applications for different measurement protocols. A RESTful API enables system communication and centralized data storage for reporting. The Posture Lab serves as an effective tool for organizations to evaluate employee postures and supports early intervention strategies, allowing timely referrals to healthcare providers if any potential musculoskeletal issues are identified. The Posture Lab has also shown medium to very high correlations with standard 2D motion analysis methods – Kinovea – for CA, SA, and KA in FHP with kyphosis measurements (r = 0.607, 0.704, and 0.992) and shown high to very high correlations in NHP with normal thoracic spine measurements (r = 0.809, 0.748, and 0.778), with significance at p < .01, utilizing the Pearson correlation coefficient.

Comprehensive housing stock information is crucial for informing the development of climate resilience strategies aiming to reduce the adverse impacts of extreme climate hazards in high-risk regions like the Caribbean. In this study, we propose an end-to-end workflow for rapidly generating critical baseline exposure data using very high-resolution drone imagery and deep learning techniques. Specifically, our work leverages the segment anything model (SAM) and convolutional neural networks (CNNs) to automate the generation of building footprints and roof classification maps. We evaluate the cross-country generalizability of the CNN models to determine how well models trained in one geographical context can be adapted to another. Finally, we discuss our initiatives for training and upskilling government staff, community mappers, and disaster responders in the use of geospatial technologies. Our work emphasizes the importance of local capacity building in the adoption of AI and Earth Observation for climate resilience in the Caribbean.

Environmental enrichment programmes are widely used to improve welfare of captive and laboratory animals, especially non-human primates. Monitoring enrichment use over time is crucial, as animals may habituate and reduce their interaction with it. In this study we aimed to monitor the interaction with enrichment items in groups of rhesus macaques (Macaca mulatta), each consisting of an average of ten individuals, living in a breeding colony. To streamline the time-intensive task of assessing enrichment programmes we automated the evaluation process by using machine learning technologies. We built two computer vision-based pipelines to evaluate monkeys’ interactions with different enrichment items: a white drum containing raisins and a non-food-based puzzle. The first pipeline analyses the usage of enrichment items in nine groups, both when it contains food and when it is empty. The second pipeline counts the number of monkeys interacting with a puzzle across twelve groups. The data derived from the two pipelines reveal that the macaques consistently express interest in the food-based white drum enrichment, even several months after its introduction. The puzzle enrichment was monitored for one month, showing a gradual decline in interaction over time. These pipelines are valuable for assessing enrichment by minimising the time spent on animal observation and data analysis; this study demonstrates that automated methods can consistently monitor macaque engagement with enrichments, systematically tracking habituation responses and long-term effectiveness. Such advancements have significant implications for enhancing animal welfare, enabling the discontinuation of ineffective enrichments and the adaptation of enrichment plans to meet the animals’ needs.

Treating images as data has become increasingly popular in political science. While existing classifiers for images reach high levels of accuracy, it is difficult to systematically assess the visual features on which they base their classification. This paper presents a two-level classification method that addresses this transparency problem. At the first stage, an image segmenter detects the objects present in the image and a feature vector is created from those objects. In the second stage, this feature vector is used as input for standard machine learning classifiers to discriminate between images. We apply this method to a new dataset of more than 140,000 images to detect which ones display political protest. This analysis demonstrates three advantages to this paper’s approach. First, identifying objects in images improves transparency by providing human-understandable labels for the objects shown on an image. Second, knowing these objects enables analysis of which distinguish protest images from non-protest ones. Third, comparing the importance of objects across countries reveals how protest behavior varies. These insights are not available using conventional computer vision classifiers and provide new opportunities for comparative research.

Fast and efficient identification is critical for reducing the likelihood of weed establishment and for appropriately managing established weeds. Traditional identification tools require either knowledge of technical morphological terminology or time-consuming image matching by the user. In recent years, deep learning computer vision models have become mature enough to enable automatic identification. The major remaining bottlenecks are the availability of a sufficient number of high-quality, reliably identified training images and the user-friendly, mobile operationalization of the technology. Here, we present the first weed identification and reporting app and website for all of Australia. It includes an image classification model covering more than 400 species of weeds and some Australian native relatives, with a focus on emerging biosecurity threats and spreading weeds that can still be eradicated or contained. It links the user to additional information provided by state and territory governments, flags species that are locally reportable or notifiable, and allows the creation of observation records in a central database. State and local weed officers can create notification profiles to be alerted of relevant weed observations in their area. We discuss the background of the WeedScan project, the approach taken in design and software development, the photo library used for training the WeedScan image classifier, the model itself and its accuracy, and technical challenges and how these were overcome.

Visual odometry (VO) is a key technology for estimating camera motion from captured images. In this paper, we propose a novel RGB-D visual odometry by constructing and matching features at the superpixel level that represents better adaptability in different environments than state-of-the-art solutions. Superpixels are content-sensitive and perform well in information aggregation. They could thus characterize the complexity of the environment. Firstly, we designed the superpixel-based feature SegPatch and its corresponding 3D representation MapPatch. By using the neighboring information, SegPatch robustly represents its distinctiveness in various environments with different texture densities. Due to the inclusion of depth measurement, the MapPatch constructs the scene structurally. Then, the distance between SegPatches is defined to characterize the regional similarity. We use the graph search method in scale space for searching and matching. As a result, the accuracy and efficiency of matching process are improved. Additionally, we minimize the reprojection error between the matched SegPatches and estimate camera poses through all these correspondences. Our proposed VO is evaluated on the TUM dataset both quantitatively and qualitatively, showing good balance to adapt to the environment under different realistic conditions.

With the rise of deep reinforcement learning (RL) methods, many complex robotic manipulation tasks are being solved. However, harnessing the full power of deep learning requires large datasets. Online RL does not suit itself readily into this paradigm due to costly and time-consuming agent-environment interaction. Therefore, many offline RL algorithms have recently been proposed to learn robotic tasks. But mainly, all such methods focus on a single-task or multitask learning, which requires retraining whenever we need to learn a new task. Continuously learning tasks without forgetting previous knowledge combined with the power of offline deep RL would allow us to scale the number of tasks by adding them one after another. This paper investigates the effectiveness of regularisation-based methods like synaptic intelligence for sequentially learning image-based robotic manipulation tasks in an offline-RL setup. We evaluate the performance of this combined framework against common challenges of sequential learning: catastrophic forgetting and forward knowledge transfer. We performed experiments with different task combinations to analyse the effect of task ordering. We also investigated the effect of the number of object configurations and the density of robot trajectories. We found that learning tasks sequentially helps in the retention of knowledge from previous tasks, thereby reducing the time required to learn a new task. Regularisation-based approaches for continuous learning, like the synaptic intelligence method, help mitigate catastrophic forgetting but have shown only limited transfer of knowledge from previous tasks.

Artificial Intelligence (AI) is reshaping the world as we know it, impacting all aspects of modern society, basically due to the advances in computer power, data availability and AI algorithms. The dairy sector is also on the move, from the exponential growth in AI research, to ready to use AI-based products, this new evolution to Dairy 4.0 represents a potential ‘game-changer’ for the dairy sector, to confront challenges regarding sustainability, welfare, and profitability. This research reflection explores the possible impact of AI, discusses the main drivers in the field and describes its origins, challenges, and opportunities. Further, we present a multidimensional vision considering factors that are not commonly considered in dairy research, such as geopolitical aspects and legal regulations that can have an impact on the application of AI in the dairy sector. This is just the beginning of the third tide of AI, and a future is still ahead. For now, the current advances in AI at on-farm level seem limited and based on the revised data, we believe that AI can be a ‘game-changer’ only if it is integrated with other components of Dairy 4.0 (such as robotics) and is fully adopted by dairy farmers.