Visual Question Answering (VQA) bridges the gap between computer vision and natural language processing, yet progress remains largely confined to high-resource languages. For low-resource languages like Urdu, research is severely hindered by the total absence of large-scale reasoning-based datasets. To address this critical gap, we introduce the first synthetic Urdu VQA dataset modeled after the CLEVR framework, specifically designed to evaluate complex, multi-step visual reasoning. We conduct a rigorous comparative analysis using both transformer-based architectures (VisualBERT, LXMERT, ViLT) and neuro-symbolic models. Our results demonstrate that the neuro-symbolic approach achieves a superior accuracy of 85.3%, outperforming the strongest transformer baseline by 7.1% while maintaining competitive processing efficiency. This work establishes a primary benchmark for Urdu VQA, demonstrating that hybrid reasoning architectures provide a robust and scalable solution for advancing multimodal AI in under-resourced linguistic contexts.

This study explores the effectiveness of transformer-based models for Named Entity Recognition (NER) in Urdu legal documents, a critical task in low-resource language processing. Given the legal texts’ specialized terminology and complex syntax, accurate entity recognition in Urdu remains challenging. We developed a legal Urdu dataset that contains 117,500 documents, generated synthetically from 47 different types of legal documents, and evaluated three BERT-based models. XLMRoBERTa, mBERT, and DistilBERT by analyzing their performance on an annotated Urdu legal dataset. mBERT demonstrated superior accuracy (0.999), and its F1 score (0.975) outperforms XLMRoBERTa and DistilBERT, highlighting its robustness in recognizing entities within low-resource languages. To ensure the privacy of the personal identifiers, all documents are anonymized. The dataset for this study is publicly hosted on Hugging Face and will be made public after the publication.

This work aims to develop an understanding of the rapidly emerging field of VideoQA, particularly in the context of instructional programming videos. It also encourages designing of system that can produce visual answer to programming based natural language questions. We introduce two datasets: CodeVidQA, with 2,104 question-answer pair links with timestamps taken from programming videos of Stack Overflow for Programming Visual Answer Localization task, and CodeVidCL with 4,331 videos (1,751 programming ,2580 non-programming) for Programming Video Classification task. In addition, we proposed a framework that adapts BigBird and SVM for video classification techniques. The proposed approach achieves a significantly high accuracy of 99.61% for video classification.