Apple will showcase a slate of computer‑vision research at this year’s IEEE/CVF Conference on Computer Vision and Pattern Recognition. The event runs June 3‑7 at the Colorado Convention Center in Denver, and Apple is both sponsor and presenter. The company plans poster and oral sessions, invited talks, a keynote, and several affinity events, underscoring its push into generative AI and multimodal models, and collaboration with partners.

Among the 20 papers Apple researchers will present are AMUSE, a benchmark for audio‑visual speaker understanding, and AToken, a unified tokenizer for vision tasks. Other highlights include DSO for bias mitigation, a large‑scale text‑guided image‑editing dataset called Pico‑Banana‑400K, and STARFlow‑V, an end‑to‑end video generator using normalizing flows. These works target core challenges across perception, alignment and efficient generation.

Colin Lea will deliver the keynote at the Generative AI for Sign Language workshop, while engineers Hsin‑Ping (Cindy) Huang and Maggie Xiao will attend the Women in Computer Vision mentorship dinner. Apple’s heavy presence at CVPR signals a strategic bet on research that can feed its on‑device AI pipelines and shape industry standards for multimodal perception for developers.