Exif2Vec: A Framework to Ascertain Untrustworthy Crowdsourced Images Using Metadata

<p indent="0mm">Volunteer geographic data, especially from social media, has emerged as a crucial resource for disaster mitigation. Particularly the included imagery can enhance situational awareness during emergencies. However, current applications of this data for disaster relief are primarily limited by a scarcity of annotated images for training, which hinders the efficacy of techniques like machine learning and deep learning, consequently restricting the data′s capacity to convey disaster information. Starting from the perspective of practical application, this paper focuses on earthquake disasters as the research subject. By integrating data acquisition, automatic deduplication, semantic annotation, and data augmentation methods, we developed a dataset of semantic segmentation in earthquake disaster detection based on social media images. This dataset mainly consists of original images collected from social media and pseudo-color images with manual semantic annotations. To ensure data quality, multiple annotators work in rotation and cross-check each other′s work. This dataset plays a significant role in improving the efficiency of social media data usage and enhancing awareness of disaster reduction efforts.

In our paper we tried to classify food images using convolutional neural network. Convolutional neural network extracts spatial features from images so it is very efficient to use convolutional neural network for image clasification problem. Recently people are sharing food images in social media and writing review on food. So there is a lot of food image in the social media but some image may not be labeled. It will be very helpful for restaurants if they can advertise their food to those people who is looking similar kind of foods they offer. Food classification system can help social media platform to identify food. Food classification system can enable an opportunity for social media platform to offer advertisement service for restaurants and beverage companies to their targeted users. It will be financially beneficial for both social media platform and beverage companies. Food classification is very difficult task because there is high variance in same category of food images. We developed a convolutional neural network model to classify food images in food-11 dataset. We also used a pre-trained Inception V3 convolutional neural network model to classify food images.

Urban land use on a building instance level is crucial geo-information for many applications yet challenging to obtain. Steet-level images are highly suited to predict building functions as the building façades provide clear hints. Social media image platforms contain billions of images, including but not limited to street perspectives. This study proposes a filtering pipeline to yield high-quality, ground-level imagery from large-scale social media image datasets to cope with this issue. The pipeline ensures all resulting images have complete and valid geotags with a compass direction to relate image content and spatial objects.We analyze our method on a culturally diverse social media dataset from Flickr with more than 28 million images from 42 cities worldwide. The obtained dataset is then evaluated in the context of a building function classification task with three classes: Commercial, residential, and other. Fine-tuned state-of-the-art architectures yield F1 scores of up to 0.51 on the filtered images. Our analysis shows that the quality of the labels from OpenStreetMap limits the performance. Human-validated labels increase the F1 score by 0.2. Therefore, we consider these labels weak and publish the resulting images from our pipeline and the depicted buildings as a weakly labeled dataset.

Social media is a routine part of every-day life for millions of people worldwide. How does engaging with social media shape enduring memories for that experience? This question is important given the popularity of certain types of content on social media platforms, such as content widely known as “fitspiration”. Two experiments involving 510 US adults (mean age = 36.82) examined memory for food and fitness-related social media images that individuals write comments about, as well as memory for other images in the context. We demonstrate that commenting on social media images boosts memory for them and weakly affects memory for conceptually related images in the same context. Exploratory analyses revealed correlations between self-reported disordered eating symptomology and effects of commenting on memory. These findings demonstrate that how people engage with social media has implications for the enduring memories of that content and may relate to behaviors and attitudes in offline lives, such as eating and body image.

Disaster response preparedness, risk mitigation, and effective post-event response are critical for national security, public health, and economic integrity. For areas impacted by extreme events, timely and accurate impact assessments of lifeline infrastructures are critical. Traditional disaster management and situational awareness platforms are not capable of handling the massive influx of available information from traditional and non-traditional sources. Next-generation analytic assessment approaches, where use of dynamically sourced, heterogeneous data with near real-time availability are integrated into a continually-updated analytics framework. The objective of this research was to develop a foundational multi-formalism modeling platform to integrate novel data sources for use in improving situational awareness after a disaster event. Specifically, to 1) provide an architecture for machine-learning based data fusion various multi-sensor datasets; 2) classify, geolocate, and temporally bin novel data sources for use in spatially-enabled machine-learning based data fusion; and 3) derive confidence-based observation data with machine-learning based prediction to fill spatial and temporal observation gaps. This project developed a reusable cloud-based pipeline architecture for ingesting, storing, and analyzing relevant data, executing a range of deep-learning models, and visualize the information geographically. The pipeline development was informed by previous PNNL investments (StreamKit) and built upon leading-edge technologies consisting of: 1) a raw datastore; 2) a data enrichment pipeline with custom and third-party analytics; 3) enriched data storage; and 4) a custom UI to display the enriched data. Machine learning/deep learning models were selected and tested to classify social media images for relevancy and presence of flood waters. Our custom trained relevance models were run on a validation dataset and the overall accuracy of the relevance classifiers ranged from 65% to 90%. We found that convolutional neural network (CNN) models were both faster and more accurate than more traditional machine learning models. We applied the relevance classifier to remove non-relevant images and then undertook a limited training of CNN models to detect flood on ~1,800 labeled images. The testing accuracy of the flood detection models were ~88%. To improve the geographic location of both social media and oblique aerial images we tested a multi-step workflow built on a pre-trained deep-learning model. To identify similarity features from test and reference scenes, we applied the Xception and Faiss models to a set of test images with inaccurate spatial locations, as well and precisely-located reference scenes from the general area. Matching reference scenes to the test scenes with the closest similarity index had limited success in improving the precision of test scene location and additional methods have been identified for testing. Data fusion between authoritative geographic data (satellite imaging) and non-authoritative crowdsourced data is an underlying theme throughout this research and methods are tested bring these data to a common geographic form through feature matching and transformations. The automated extraction of supporting information for disaster response from oblique aerial images or ground-level social media images is accomplished through semantic image segmentation and transformation. The non-authoritative data are used to validate the authoritative data collections and processing and optionally, t+0 flood models. In addition, the common form geographic data are combined with other static landscape features to develop spatially-explicit training vectors used in deep generative statistical inference models to develop complex non-linear relationships in static and dynamic data and ultimately derive a spatially continuous and probabilistic assessment of flood damage with the event domain.

Floods are among the most frequent and catastrophic natural disasters and affect millions of people worldwide. It is important to create accurate flood maps to plan (offline) and conduct (real-time) flood mitigation and flood rescue operations. Arguably, images collected from social media can provide useful information for that task, which would otherwise be unavailable. We introduce a computer vision system that estimates water depth from social media images taken during flooding events, in order to build flood maps in (near) real-time. We propose a multi-task (deep) learning approach, where a model is trained using both a regression and a pairwise ranking loss. Our approach is motivated by the observation that a main bottleneck for image-based flood level estimation is training data: it is difficult and requires a lot of effort to annotate uncontrolled images with the correct water depth. We demonstrate how to efficiently learn a predictor from a small set of annotated water levels and a larger set of weaker annotations that only indicate in which of two images the water level is higher, and are much easier to obtain. Moreover, we provide a new dataset, named DeepFlood, with 8145 annotated ground-level images, and show that the proposed multi-task approach can predict the water level from a single, crowd-sourced image with ≈11 cm root mean square error.

The popularity of social media has grown considerably over the past few years and has taken a foothold in society. This study examined the connection between extensive social media use and the way people view their bodies in relation to what is seen in the images and messages portrayed on social media. Emphasis is placed on social media and whether it has a negative impact on body image and if it leads to body dissatisfaction. A mixed method approach was taken to collect data on females residing in Tobago and how they perceive body image. Survey participation was both voluntary and anonymous. Participants were females above the age of 15. Additionally, focus groups were conducted among Tobago women ranging between ages 16- 56 to further supplement the statistical data collected. The study and the data collected examined how social media influenced the way that women view themselves. The data collected and the research took into consideration how the participants were affected by their exposure to the various social media platforms and their messages. While the information gathered shows there is a correlation between social media and negative body image, there is no evidence to prove that social media has a direct impact on one’s body image.

This article examines how political leaders use social media to shape their image during elections. Social media and digital platforms have changed political communication, allowing politicians to craft their image carefully. Social media has changed politics, yet misinformation and public opinion manipulation are concerns. Politicians' online avatars may misrepresent their goals and policies, overshadowing public policy discussions. This study uses a comprehensive analysis of the literature from the Scopus database to analyze how political elites shape their social media image during elections in different nations. This research examines how political leaders use social media to shape their public persona during election campaigns in several countries. The results will reveal political elites' social media image-building strategies during elections, including personalized methods, visual cues' impact on engagement, and preferred platforms for visual communication. This extensive study of the long-term consequences of visual communication tactics shows that politicians’ use of social media to convey stories has a lasting impact on public perception and engagement in politics.

Purpose/Rationale The study examines how consumers look at sponsorship messaging embedded within Instagram images on the International Volleyball Federation’s official beach volleyball account, investigating how social media consumers dedicate visual attention to the photographs. Design The study followed an experimental design using eye-tracking technology with a within-subjects protocol. Researchers recorded the gaze behavior of participants (n = 41) on social media images (n = 18) featuring beach volleyball athletes who played on the 2023 Volleyball World Beach Pro Tour Elite16. Findings The study’s results revealed that consumers dedicated most of their visual attention to the athletes in the images compared to the sponsorship messaging. In addition, participants were less likely to look at sponsors when the athletes were in defensive positions compared to offensive positions or posed positions. Practical implications The lack of visual attention to sponsorship messaging reinforces the need for social media marketers to be strategic in how they select images for Instagram and incorporate sponsors into their social media feeds. Research contribution Results highlight the need to better understand how consumers dedicate visual attention to sponsors embedded within social media posts. Originality/Value This article widens the methodological diversity in sports and social media research by using eye-tracking technology.

Promises and pitfalls of using computer vision to make inferences about landscape preferences: Evidence from an urban-proximate park system

This paper presents a multiple feature fusion method using topic model for social image visualization. Images in social media are represented from several aspects such as their visual information and tags. The proposed method extracts low-level features from social images and their tags and calculates their integrated high-level features. Specifically, the proposed method applies multilayer multimodal probabilistic Latent Semantic Analysis (mm-pLSA) to the low-level visual and tag features to obtain the high-level features. Then, by applying dimensionality reduction techniques to the obtained features, successful visualization becomes feasible.

In an article in this issue of Sex Roles, I proposed a transactional model of social media effects on body image concerns that emphasizes the role played by individual vulnerability characteristics, gratifications sought from social media, and a host of mediating psychological processes. Commentators offered thoughtful responses to the model, pointing to strengths, but also targeting a number of areas for additional emphasis and ameliorative attention. This article, with a focus on the U.S. context, reviews each of the papers. It summarizes their main suggestions, and pulls together the host of forward-looking ideas, including the role played by cultural forces, the interplay between social and conventional mass media, and an emphasis on the salutary effects of social media on body image processes.

Accurately estimating urban floodwater depth is a critical step in enhancing urban resilience and strengthening disaster prevention and mitigation capabilities. Traditional methods relying on hydrological monitoring stations and numerical simulations suffer from limitations such as sparse spatial coverage, insufficient validation data, limited accuracy, and delayed fast performance. In contrast, social media data—characterized by its vast volume and fast availability, can effectively compensate for these shortcomings. When processed using artificial intelligence (AI) algorithms, such data can significantly improve credibility, disaster perception speed, and water depth estimation accuracy. To address these challenges, this paper proposes a robust and widely applicable method for rapid urban flood depth perception. The approach integrates AI technology and social media data to construct an AI framework capable of perceiving urban physical parameters through multimodal big data fusion without costly model training. By leveraging the near real-time and widespread nature of social media, an automated web crawler collects flood images and their textual descriptions (including reference objects), eliminating the need for additional hardware investments. The framework uses predefined prompts and pre-trained models to automatically perform relevance verification, duplicate filtering, object detection, and feature extraction, requiring no manual data annotation or model training. With only a minimal amount of water depth annotated data and compressed cross-modal feature vectors as training input, a lightweight Multilayer Perceptron (MLP) achieves high-precision depth estimation based on reference objects. This method avoids the need for large-scale model fine-tuning, allowing rapid training even on devices without GPUs. Experiments demonstrate that the proposed method reduces the Mean Square Error (MSE) by over 80%, processes each image in less than 0.5 s (more than 20 times faster than existing large-model approaches), and exhibits strong robustness to changes in perspective and image quality. The solution is fully compatible with existing infrastructure such as surveillance cameras, offering an efficient and reliable approach for fast flood monitoring in urban hydrology and water engineering applications.

Planners, policy makers, and scholars are increasingly using social media data to study public life in cities. Yet, such projects tend to be limited by three commitments that shape the imaginaries of such data-driven urbanism, namely 1) bias towards textual social media; 2) fixation on geotag-ontologies; and 3) seeing the subjective nature of data as a bias. The consequence is that the potential of digital traces for renewing the empirical ground of digital urban studies is not fully realized. To open alternative imaginaries around data-driven urbanism, we provide a bibliometric review of these trends and suggest that social media images could be used to study place attachments and explore how people experience cities, bridging ethnographic research questions with the computational agenda. Second, to exemplify what can be gained from such a re-orientation of urban projects, we deliver a digital methods study of 39K Instagram posts from 2020 and explore how people in Denmark attached to different environments during the first nine months of COVID-19. The case demonstrates that we might open new empirical routes in urban studies by centering image data, moving beyond geotag-ontologies, and foregrounding the subjectivity of data as an analytical opportunity, rather than a problem.

Social media posts consist of a large number of flood-related images and data. These images can be helpful for the flood image classification process and necessary to produce real-time information about road accessibility during rescue and disaster management. In this paper, social media images are classified into flooded and non-flooded categories using EfficientNet variants with MediaEval 2017 Disaster Image Retrieval from the social media dataset. The transfer learning technique is employed to achieve good performance and efficiency. Then different variants are compared using various metrics after and before freezing. The higher variants are intentionally avoided, as it does not perform with better efficiency and accuracy. The B3 variant outperforms all other variance when considering model validation accuracy and false positive rate after unfreezing top layers. But when analysing the classification report, all variants show almost equal performance and weighted average F1-score.KeywordsEfficientNetFloodSocial media