Researchers have introduced BloomBench, a new cognitively human-grounded, bilingual (English-Arabic) multimodal benchmark for Vision-Language Models (VLMs), as part of the Almieyar benchmarking series. Grounded in Bloom's Taxonomy, it systematically evaluates six levels of cognition—Remember, Understand, Apply, Analyze, Evaluate, Create—through carefully designed image-question-answer tasks. A comprehensive study using BloomBench revealed that state-of-the-art VLMs exhibit strong semantic understanding but struggle significantly with factual recall and creative synthesis, alongside a critical performance gap between Arabic and English. Why it matters: This benchmark provides a crucial tool for diagnosing cognitive weaknesses in current VLMs and lays the groundwork for developing more cognitively aligned and inclusive multimodal AI, particularly for cross-lingual applications.
The Technology Innovation Institute (TII) in Abu Dhabi has launched Falcon 2, a new series of large language models including the Falcon 2 11B and Falcon 2 11B VLM. The Falcon 2 11B outperforms Meta’s Llama 3 (8B) and performs on par with Google’s Gemma 7B, as verified by Hugging Face. Falcon 2 11B VLM is TII's first multimodal model with vision-to-language capabilities and is open-source, making it accessible to developers. Why it matters: This release strengthens the UAE's position in AI research and development, providing open-source models that can be deployed on smaller infrastructures and used in diverse sectors.
A new study compares vision-language models (VLMs) to YOLOv8 for wastewater treatment plant (WWTP) identification in satellite imagery across the MENA region. VLMs like Gemma-3 demonstrate superior zero-shot performance compared to YOLOv8, trained on a dataset of 83,566 satellite images from Egypt, Saudi Arabia, and UAE. The research suggests VLMs offer a scalable, annotation-free alternative for remote sensing of WWTPs.
Researchers introduce MATRIX, a vision-centric agent tuning framework for robust tool-use reasoning in VLMs. The framework includes M-TRACE, a dataset of 28.5K multimodal tasks with 177K verified trajectories, and Pref-X, a set of 11K automatically generated preference pairs. Experiments show MATRIX consistently outperforms open- and closed-source VLMs across three benchmarks.
This paper introduces MOTOR, a multimodal retrieval and re-ranking approach for medical visual question answering (MedVQA) that uses grounded captions and optimal transport to capture relationships between queries and retrieved context, leveraging both textual and visual information. MOTOR identifies clinically relevant contexts to augment VLM input, achieving higher accuracy on MedVQA datasets. Empirical analysis shows MOTOR outperforms state-of-the-art methods by an average of 6.45%.
MBZUAI researchers introduce UniMed-CLIP, a unified Vision-Language Model (VLM) for diverse medical imaging modalities, trained on the new large-scale, open-source UniMed dataset. UniMed comprises over 5.3 million image-text pairs across six modalities: X-ray, CT, MRI, Ultrasound, Pathology, and Fundus, created using LLMs to transform classification datasets into image-text formats. UniMed-CLIP significantly outperforms existing generalist VLMs and matches modality-specific medical VLMs in zero-shot evaluations, improving over BiomedCLIP by +12.61 on average across 21 datasets while using 3x less training data.
Researchers at MBZUAI have developed GeoChat, a new vision-language model (VLM) specifically designed for remote sensing imagery. GeoChat addresses the limitations of general-domain VLMs in accurately interpreting high-resolution remote sensing data, offering both image-level and region-specific dialogue capabilities. The model is trained on a novel remote sensing multimodal instruction-following dataset and demonstrates strong zero-shot performance across tasks like image captioning and visual question answering.
Researchers from MBZUAI, IBM, and ServiceNow introduced GEOBench-VLM, a benchmark for evaluating vision-language models on Earth observation tasks using satellite and aerial imagery. The benchmark includes over 10,000 human-verified instructions across 31 sub-tasks spanning object classification, localization, change detection, and more. GEOBench-VLM addresses the gap in current VLMs' ability to perform spatially grounded reasoning and change detection in satellite imagery. Why it matters: This benchmark will drive progress in AI's ability to analyze satellite data for critical applications like disaster response, climate monitoring, and urban planning in the Middle East and globally.