ArtScape Wins a Best Paper Honorable Mention at UIST 2026
Congratulations to Lucy Jiang and Jared Hwang, whose UIST 2026 paper ArtScape: Detecting, Characterizing, and Interactively Exploring Urban Art at Scale Using Vision Language Models received a Best Paper Honorable Mention, awarded to the top 5% of submissions! It is Jared's first paper award in the lab!
Murals, mosaics, sculptures, and graffiti make our cities vibrant, but they are hard to discover, especially for blind and low-vision people. ArtScape uses a vision language model to find and describe artwork in street view imagery of Seattle, then presents it on an interactive map so people can explore art across neighborhoods. One of our favorite findings: even when the VLM "curator" got something wrong, it sometimes got participants to engage more deeply with the artwork. Lucy and Jared co-presented the paper in the Interactive Data Exploration session, chaired by our own Chu Li. Lucy also presented an ArtScape demo at ASSETS 2026 in Porto the week before. Watch the video.