Apple Details ICCV 2025 Lineup With Eight Papers, Workshop Keynote, and Honolulu Exhibit

Overview

• Apple confirmed participation at ICCV 2025 in Honolulu from October 19–23 with eight accepted papers spanning poster and oral sessions.
• “Scaling Laws for Native Multimodal Models” is scheduled as an oral presentation on Tuesday, October 21 from 8:45–10:00 a.m. in Oral Session 1A (Exhibit Hall III), followed by a poster from 11:30 a.m.–1:30 p.m. at #227.
• On Sunday, October 19, Apple’s C. Thomas delivers a VISION workshop keynote from 9:15–9:45 a.m. at the Hawai’i Convention Center, and Apple researchers serve as mentors at the WiCV mentorship dinner following the 1:00–5:00 p.m. workshop.
• Apple will staff booth #220 at the Honolulu Convention Center during exhibition hours: Tuesday, October 21 (11:30 a.m.–5:00 p.m.), Wednesday, October 22 (10:45 a.m.–4:30 p.m.), and Thursday, October 23 (10:45 a.m.–4:30 p.m.) HST.
• The papers cover evaluation and generative themes including ETVA for text-to-video alignment, MM-Spatial for 3D understanding in multimodal LLMs, STIV and UniVG for image/video generation, visual in-context learning with Stable Diffusion, open-world segmentation, and UINavBench for agent evaluation.