Search
Apple Unveils Multimodal AI Model UniGen 1.5, Marking a Leap Forward in Image Processing Technology
2025-12-19 /
Read about 0 minute
Author：小编
Apple's research squad has just rolled out the multimodal AI model, UniGen 1.5. This innovative model, for the very first time, seamlessly blends three core functionalities—image comprehension, creation, and modification—within a single, cohesive framework. Through the implementation of a coordinated training phase that incorporates editing directives and a unified reward mechanism, the model effectively tackles challenges such as imprecise interpretation of image editing commands and inconsistent quality benchmarks stemming from the diverse array of tasks. In rigorous benchmark assessments like GenEval, DPG-Bench, and ImgEdit, the model showcased exceptional prowess, rivaling the capabilities of proprietary models such as GPT-Image-1.
-
C114 Communication Network -
Communication Home
7 X 24 Track global technological trends
