Reproduce “Make-A-Story: Visual Memory Conditioned Consistent Story Generation” 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
Tanzila Rahman, Hsin-Ying Lee, Jian Ren, Sergey Tulyakov, Shweta Mahajan, Leonid Sigal
Story-LDM generates story illustrations from sentence sequences by conditioning each frame on its sentence and earlier text and visual frames through autoregressive memory-attention. On reference-bearing MUGEN text, its character accuracy is 93.40, compared with 31.39 for LDM. The result supports the full framework on this test setting; on Flintstones, background accuracy falls versus the reference-bearing LDM baseline.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.