![Illustration of a multi-story indoor house layout used in the CapNav benchmark. The diagram asks: “Can [AGENT] go from the basement foyer to the top floor rest area?” The basement foyer (START) is at the bottom left, and the top floor rest area (END) is at the top right. Colored paths show different possible routes through labeled areas including Garage, First Floor Hallway, and Elevator. Icons on the left represent four agent types: a wheelchair user (allowed, green check), a quadruped robot (allowed, green check), a human (allowed, green check), and a sweeping robot (not allowed, red X). Red X markers and warning labels indicate constraints such as “Cannot go up stairs” and “Cannot operate elevator,” showing how different mobility capabilities affect route feasibility.](/media/projects/images/CapNav-Thumbnail.png.500x300_q85_box-135%2C10%2C2826%2C1627_crop_upscale.png)
CapNav
2025 – Present
Xia Su
PhD Student, Computer Science
University of Washington
Xia Su was a PhD Student in the Makeability Lab for 6.1 years (Apr 2020 to Aug 2026). They contributed to 9 projects, including Accessibility Scout, CapNav, and DIAM: Drone-based Indoor Accessibility Mapping, as well as 14 publications. During their time in the lab, Xia mentored 3 Makeability Lab students: Qiaochu (Steve) Liu, Richard Weiye Zhang, and Jackson (Jun) Ma.
Projects (8/9)
DepthScape
2024 – Present
RAIS: Robotic Accessibility Indoor Scanner
2023 – Present
DIAM: Drone-based Indoor Accessibility Mapping
2024 – Present
RASSAR
2022 – 2024
FlyMeThrough
2024 – Present
Accessibility Scout
2024 – Present
Papers (6/14)
CapNav: Benchmarking Vision Language Models on Capability-conditioned Indoor Navigation
Proceedings of CVPR 2026 | Acceptance Rate: 25.4% (4090 / 16092)
“Does the cafe entrance look accessible? Where is the door?” Towards Geospatial AI Agents for Visual Inquiries
Workshop Proceedings of ICCV CV4A11y Workshop 2025
FlyMeThrough: Human-AI Collaborative 3D Indoor Mapping with Commodity Drones
Proceedings of UIST 2025 | Acceptance Rate: 22.2% (210 / 947)
Accessibility Scout: Personalized Accessibility Scans of Built Environments
Proceedings of UIST 2025 | Acceptance Rate: 22.2% (210 / 947)
Authoring 2.5D Designs with Depth Estimation
Extended Abstract Proceedings of CHI 2025 | Acceptance Rate: 32.8% (620 / 1888)
Videos (1/1)
Talks (3/3)
FlyMeThrough: Human-AI Collaborative 3D Indoor Mapping with Commodity Drones
Sep 30, 2025 | UIST
Busan, South Korea
RASSAR: Room Accessibility and Safety Scanning in Augmented Reality
May 15, 2024 | CHI
Honolulu, Hawaiʻi