VGAP — the Computer Vision, Generative AI & Perception Lab at HBKU
We work on problems we find genuinely interesting: diffusion models, multi-modal LLMs, 3D vision, and uncertainty-aware perception. We also like taking that research into the real world — from fire-safety compliance to smart cities and Arabic generative AI, with government and industry partners across Qatar, the GCC, and the wider Middle East (MENA).
Peer-reviewed work in Computer Vision & Generative AI
Publications at NeurIPS, ICCV, CVPR, SIGGRAPH and ICLR spanning diffusion models, multi-modal LLMs, 3D vision, and uncertainty estimation.
Browse publications →SolutionsSee how our research can solve your real-life problems
From fire-safety compliance to telecom AI assistants and smart-city traffic monitoring — see how our research translates into deployed, working systems.
See our solutions →Solutions by sector
View all →
Featured publications
View all →Mind-the-Glitch: Visual Correspondence for Detecting Inconsistencies in Subject-Driven Generation
Introduces a method for detecting visual inconsistencies in subject-driven image generation by leveraging visual correspondence, improving reliabil…
EditCLIP: Representation Learning for Image Editing
A representation-learning approach tailored to image editing, learning embeddings that capture the semantics of an edit itself rather than just ima…
PartEdit: Fine-Grained Image Editing using Pre-Trained Diffusion Models
A method for precise, part-level image editing built on pre-trained diffusion models, allowing targeted edits to specific object parts without dist…
Latest news
View all →VGAP Lab launches at HBKU
We're launching VGAP — the Computer Vision, Generative AI & Perception Lab at HBKU — bringing together academic research and applied AI work for government and industry partners across Qatar and the GCC.
"Mind-the-Glitch" accepted as a Spotlight at NeurIPS 2025
Our paper on detecting visual inconsistencies in subject-driven generation was accepted as a Spotlight presentation at NeurIPS 2025.
Four papers accepted at ICCV, SIGGRAPH & ICLR 2025
EditCLIP, PlaceIt3D and ZeroKey were accepted at ICCV 2025, PartEdit at SIGGRAPH 2025, and Build-A-Scene at ICLR 2025 — a strong showing across generative AI and 3D vision venues.