ResearchMedium importance
Open Vision Banana, a fully open recipe for turning image generators into vision systems, tops the votes on Hugging Face Papers
The paper starts from the idea that pretrained image generators understand the real world beyond drawing images. Vision Banana already exploited this by treating vision tasks as conditional image editing, but without publishing data, weights, code or details. Open Vision Banana (OVB) is an open, reproducible recipe for converting these generators into multi-task perception systems, and its analyses point to what matters most for generalization: color encoding, prompt design, how parameters are adapted and the balance between tasks. On the morning of 11/14 it was the most-voted paper on the Hugging Face Papers list, with 32 votes.
Published in edition no. 29 of Noticiário do Joãozinho ·
Translated by Joãozinho AI from the original Brazilian Portuguese edition. Facts, numbers and sources are the same as in the original; quotes, video titles and source names stay in their original language. Read the original in Portuguese
Original source
arXiv / Hugging Face PapersPage on Hugging Face Papers
https://arxiv.org/abs/2611.15844