Abu Dhabi has taken a decisive step in global artificial intelligence (AI) with Falcon Perception, a multimodal model that enables machines to efficiently see, read and interpret the physical world.
Developed by the Technology Innovation Institute (TII), the applied research arm of the emirate’s Advanced Technology Research Council, Falcon Perception expands the UAE’s AI ecosystem by combining visual and linguistic capabilities.
As the global AI race intensifies, the UAE is positioning itself among countries capable of deploying advanced multimodal systems at scale, and Falcon Perception is at the heart of this ambition. With approximately 600 million parameters, Falcon Perception is significantly more compact than many prominent multimodal models, which often use billions of parameters.
“Our goal with Falcon Perception was to challenge the common assumption that vision systems need to rely on complex multi-level architectures. By demonstrating that a single, dense transformer can efficiently handle perceptual tasks, we open the door to a new generation of scalable multimodal systems,” said Hakim Hasid, principal investigator at the TII Artificial Intelligence and Digital Research Center.
This balance between efficiency and performance is indicative of a broader trend in AI. Rather than increasing the number of parameters or requiring large computing resources, researchers focus on optimizing model designs, such as efficient transformer variants, to achieve superior results even on resource-constrained hardware.
Multimodal AI is widely recognized as the next frontier in artificial intelligence. Large-scale language models (LLMs) have dominated recent advances, but as AI expands into robotics, manufacturing, and intelligent infrastructure, the ability of machines to interpret and interact with the physical world becomes important.
Falcon Perception employs a unified transformer-based architecture that enables end-to-end integration of visual and linguistic features at the model input level. Unlike traditional pipelines that combine separately trained computer vision and NLP modules, Falcon Perception performs processing and inference across modalities directly within a shared network, reducing inference latency and deployment complexity.
As a result, the system uses natural language prompts to interpret complex multi-object visual scenes. Users can instruct the model to identify, count, and segment specific objects in an image, and Falcon Perception returns bounding boxes, segmentation masks, or text output, even in crowded and complex environments.
Such capabilities have clear implications for the industry. In manufacturing, this model could enable automated inspection and defect detection. Robotics enables machines to follow natural language instructions in dynamic environments. In enterprise environments, it can streamline large-scale document processing and visual data labeling.
For TII, this launch is not only a technological milestone, but also a step in a broader national strategy. Since beginning its AI journey, the UAE has prioritized building sovereign capacity, ensuring domestic development, responsible governance, and aligning critical technologies with long-term economic goals.
“From the beginning, Abu Dhabi has focused on advancing AI research that is secure, transparent and sovereignly designed,” said Ray O Johnson, CEO of TII. “Our goal is to enable governments, industry and society to deploy AI with confidence.”
TII’s work spans AI safety, assessment and deployment frameworks, and large-scale research programs. The main outcome of this effort is Falcon, the UAE’s indigenous LLM, which was first launched by TII in 2023. Falcon quickly gained international attention for its performance and for being released as an open source model. This reflects Abu Dhabi’s belief that openness and governance can coexist.
Falcon is positioned as part of a broader national AI development system, not just a technological achievement. By combining scientific research with agile decision-making at the government level, Abu Dhabi aims to accelerate adoption while maintaining oversight and trust.
