Crowdsource, Crawl, or Generate? Creating SEA-VL, a Multicultural Vision-Language Dataset for Southeast Asia

Southeast Asia (SEA) is a region of extraordinary linguistic and cultural diversity, yet it remains significantly underrepresented in vision-language (VL) research. This often results in artificial intelligence (AI) models that fail to capture SEA cultural nuances. To fill this gap, we present SEA-VL, an open-source initiative dedicated to developing high-quality, culturally relevant data for SEA languages. By involving contributors from SEA countries, SEA-VL aims to ensure better cultural relevance and diversity, fostering greater inclusivity of underrepresented languages in VL research. Beyond crowdsourcing, our initiative goes one step further in the exploration of the automatic collection of culturally relevant images through crawling and image generation. First, we find that image crawling achieves approximately ~85% cultural relevance while being more cost- and time-efficient than crowdsourcing. Second, despite the substantial progress in generative vision models, synthetic images remain unreliable in accurately reflecting SEA cultures. The generated images often fail to reflect the nuanced traditions and cultural contexts of the region. Collectively, we gather 1.28M SEA culturally-relevant images, more than 50 times larger than other existing datasets. Through SEA-VL, we aim to bridge the representation gap in SEA, fostering the development of more inclusive AI systems that authentically represent diverse cultures across SEA.

Improving Clinical Question Answering with Multi-Task Learning: A Joint Approach for Answer Extraction and Medical Categorization

Clinical Question Answering (CQA) plays a crucial role in medical decision-making, enabling physicians to extract relevant information from Electronic Medical Records (EMRs). While transformer-based models such as BERT, BioBERT, and ClinicalBERT have demonstrated state-of-the-art performance in CQA, existing models lack the ability to categorize extracted answers, which is critical for structured retrieval, content filtering, and medical decision support. To address this limitation, we introduce a Multi-Task Learning (MTL) framework that jointly trains CQA models for both answer extraction and medical categorization. In addition to predicting answer spans, our model classifies responses into five standardized medical categories: Diagnosis, Medication, Symptoms, Procedure, and Lab Reports. This categorization enables more structured and interpretable outputs, making clinical QA models more useful in real-world healthcare settings. We evaluate our approach on emrQA, a large-scale dataset for medical question answering. Results show that MTL improves F1-score by 2.2% compared to standard fine-tuning, while achieving 90.7% accuracy in answer categorization. These findings suggest that MTL not only enhances CQA performance but also introduces an effective mechanism for categorization and structured medical information retrieval.

MVTamperBench: Evaluating Robustness of Vision-Language Models

Recent advancements in Vision-Language Models (VLMs) have enabled significant progress in complex video understanding tasks. However, their robustness to real-world manipulations remains underexplored, limiting their reliability in critical applications. To address this gap, we introduce MVTamperBench, a comprehensive benchmark designed to evaluate VLM's resilience to video tampering effects, including rotation, dropping, masking, substitution, and repetition. By systematically assessing state-of-the-art models, MVTamperBench reveals substantial variability in robustness, with models like InternVL2-8B achieving high performance, while others, such as Llama-VILA1.5-8B, exhibit severe vulnerabilities. To foster broader adoption and reproducibility, MVTamperBench is integrated into VLMEvalKit, a modular evaluation toolkit, enabling streamlined testing and facilitating advancements in model robustness. Our benchmark represents a critical step towards developing tamper-resilient VLMs, ensuring their dependability in real-world scenarios. Project Page: https://amitbcp.github.io/MVTamperBench/

Survey of Large Multimodal Model Datasets, Application Categories and Taxonomy

Multimodal learning, a rapidly evolving field in artificial intelligence, seeks to construct more versatile and robust systems by integrating and analyzing diverse types of data, including text, images, audio, and video. Inspired by the human ability to assimilate information through many senses, this method enables applications such as text-to-video conversion, visual question answering, and image captioning. Recent developments in datasets that support multimodal language models (MLLMs) are highlighted in this overview. Large-scale multimodal datasets are essential because they allow for thorough testing and training of these models. With an emphasis on their contributions to the discipline, the study examines a variety of datasets, including those for training, domain-specific tasks, and real-world applications. It also emphasizes how crucial benchmark datasets are for assessing models' performance in a range of scenarios, scalability, and applicability. Since multimodal learning is always changing, overcoming these obstacles will help AI research and applications reach new heights.

Enhancing Document AI Data Generation Through Graph-Based Synthetic Layouts

The development of robust Document AI models has been constrained by limited access to high-quality, labeled datasets, primarily due to data privacy concerns, scarcity, and the high cost of manual annotation. Traditional methods of synthetic data generation, such as text and image augmentation, have proven effective for increasing data diversity but often fail to capture the complex layout structures present in real world documents. This paper proposes a novel approach to synthetic document layout generation using Graph Neural Networks (GNNs). By representing document elements (e.g., text blocks, images, tables) as nodes in a graph and their spatial relationships as edges, GNNs are trained to generate realistic and diverse document layouts. This method leverages graph-based learning to ensure structural coherence and semantic consistency, addressing the limitations of traditional augmentation techniques. The proposed framework is evaluated on tasks such as document classification, named entity recognition (NER), and information extraction, demonstrating significant performance improvements. Furthermore, we address the computational challenges of GNN based synthetic data generation and propose solutions to mitigate domain adaptation issues between synthetic and real-world datasets. Our experimental results show that graph-augmented document layouts outperform existing augmentation techniques, offering a scalable and flexible solution for training Document AI models.

LLM for Barcodes: Generating Diverse Synthetic Data for Identity Documents

Accurate barcode detection and decoding in Identity documents is crucial for applications like security, healthcare, and education, where reliable data extraction and verification are essential. However, building robust detection models is challenging due to the lack of diverse, realistic datasets an issue often tied to privacy concerns and the wide variety of document formats. Traditional tools like Faker rely on predefined templates, making them less effective for capturing the complexity of real-world identity documents. In this paper, we introduce a new approach to synthetic data generation that uses LLMs to create contextually rich and realistic data without relying on predefined field. Using the vast knowledge LLMs have about different documents and content, our method creates data that reflects the variety found in real identity documents. This data is then encoded into barcode and overlayed on templates for documents such as Driver's licenses, Insurance cards, Student IDs. Our approach simplifies the process of dataset creation, eliminating the need for extensive domain knowledge or predefined fields. Compared to traditional methods like Faker, data generated by LLM demonstrates greater diversity and contextual relevance, leading to improved performance in barcode detection models. This scalable, privacy-first solution is a big step forward in advancing machine learning for automated document processing and identity verification.

TLDR: Traffic Light Detection using Fourier Domain Adaptation in Hostile WeatheR

The scarcity of comprehensive datasets in the traffic light detection and recognition domain and the poor performance of state-of-the-art models under hostile weather conditions present significant challenges. To address these issues, this paper proposes a novel approach by merging two widely used datasets, LISA and S2TLD. The merged dataset is further processed to tackle class imbalance, a common problem in this domain. This merged dataset becomes our source domain. Synthetic rain and fog are added to the dataset to create our target domain. We employ Fourier Domain Adaptation (FDA) to create a final dataset with a minimized domain gap between the two datasets, helping the model trained on this final dataset adapt to rainy and foggy weather conditions. Additionally, we explore Semi-Supervised Learning (SSL) techniques to leverage the available data more effectively. Experimental results demonstrate that models trained on FDA-augmented images outperform those trained without FDA across confidence-dependent and independent metrics, like mAP50, mAP50-95, Precision, and Recall. The best-performing model, YOLOv8, achieved a Precision increase of 5.1860%, Recall increase of 14.8009%, mAP50 increase of 9.5074%, and mAP50-95 increase of 19.5035%. On average, percentage increases of 7.6892% in Precision, 19.9069% in Recall, 15.8506% in mAP50, and 23.8099% in mAP50-95 were observed across all models, highlighting the effectiveness of FDA in mitigating the impact of adverse weather conditions on model performance. These improvements pave the way for real-world applications where reliable performance in challenging environmental conditions is critical.

Snowy Scenes,Clear Detections: A Robust Model for Traffic Light Detection in Adverse Weather Conditions

With the rise of autonomous vehicles and advanced driver-assistance systems (ADAS), ensuring reliable object detection in all weather conditions is crucial for safety and efficiency. Adverse weather like snow, rain, and fog presents major challenges for current detection systems, often resulting in failures and potential safety risks. This paper introduces a novel framework and pipeline designed to improve object detection under such conditions, focusing on traffic signal detection where traditional methods often fail due to domain shifts caused by adverse weather. We provide a comprehensive analysis of the limitations of existing techniques. Our proposed pipeline significantly enhances detection accuracy in snow, rain, and fog. Results show a 40.8% improvement in average IoU and F1 scores compared to naive fine-tuning and a 22.4% performance increase in domain shift scenarios, such as training on artificial snow and testing on rain images.

Unveiling Themes in Judicial Proceedings: A Cross-Country Study Using Topic Modeling on Legal Documents from India and the UK

Legal documents are indispensable in every country for legal practices and serve as the primary source of information regarding previous cases and employed statutes. In today's world, with an increasing number of judicial cases, it is crucial to systematically categorize past cases into subgroups, which can then be utilized for upcoming cases and practices. Our primary focus in this endeavor was to annotate cases using topic modeling algorithms such as Latent Dirichlet Allocation, Non-Negative Matrix Factorization, and Bertopic for a collection of lengthy legal documents from India and the UK. This step is crucial for distinguishing the generated labels between the two countries, highlighting the differences in the types of cases that arise in each jurisdiction. Furthermore, an analysis of the timeline of cases from India was conducted to discern the evolution of dominant topics over the years.

IRLCov19: A Large COVID-19 Multilingual Twitter Dataset of Indian Regional Languages

Emerged in Wuhan city of China in December 2019, COVID-19 continues to spread rapidly across the world despite authorities having made available a number of vaccines. While the coronavirus has been around for a significant period of time, people and authorities still feel the need for awareness due to the mutating nature of the virus and therefore varying symptoms and prevention strategies. People and authorities resort to social media platforms the most to share awareness information and voice out their opinions due to their massive outreach in spreading the word in practically no time. People use a number of languages to communicate over social media platforms based on their familiarity, language outreach, and availability on social media platforms. The entire world has been hit by the coronavirus and India is the second worst-hit country in terms of the number of active coronavirus cases. India, being a multilingual country, offers a great opportunity to study the outreach of various languages that have been actively used across social media platforms. In this study, we aim to study the dataset related to COVID-19 collected in the period between February 2020 to July 2020 specifically for regional languages in India. This could be helpful for the Government of India, various state governments, NGOs, researchers, and policymakers in studying different issues related to the pandemic. We found that English has been the mode of communication in over 64% of tweets while as many as twelve regional languages in India account for approximately 4.77% of tweets.