Reproduce “DiffAuction: An Auction-Driven Framework for Generative Image Advertising” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Yijie Su, Ruiyang Si, Tonghan Wang
DiffAuction allocates product exposure within a jointly generated advertising image. It calibrates product-specific residuals to requested shares and adds share-conditioned viewpoint prompts, then scores images with a content-based click-intent proxy for an auction. Across 200 scenes and 19 share settings, residual calibration raises Spearman share–CTI correlation from 0.7837 to 0.8346; this is not evidence of actual clicks.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.