Text to Figure文生图poster

DSPO Score-Matching Alignment for Diffusion Models

A conference poster presenting DSPO, a score-matching-based alignment algorithm for diffusion models. It compares DSPO against baselines like Diffusion-DPO using win-rate charts and performance tables on Stable Diffusion 1.5 and XL across datasets like PickV2 and HPSV2.

论文上下文

Paper title: DSPO: Direct Score Preference Optimization for Diffusion Model Alignment Abstract: A conference poster presenting DSPO, a score-matching-based alignment algorithm for diffusion models. It compares DSPO against baselines like Diffusion-DPO using win-rate charts and performance tables on Stable Diffusion 1.5 and XL across datasets like PickV2 and HPSV2. Paper body (method & results): <!DOCTYPE html> <html lang="en"> <head> <meta content="text/html; charset=utf-8" http-equiv="content-type"/> <title>Divergence Minimization Preference Optimization for Diffusion Model Alignment</title>  <meta content="width=device-width, initial-scale=1, shrink-to-fit=no" name="viewport"/> <link href="https://cdn.jsdelivr.net/npm/bootstrap@5.3.0/dist/css/bootstrap.min.css" rel="stylesheet" type="text/css"/> <link href="/static/browse/0.3.4/css/ar5iv.0.7.9.min.css" rel="stylesheet" type="text/css"/> <link href="/static/browse/0.3.4/css/ar5iv-fonts.0.7.9.min.css" rel="stylesheet" type="text/css"/> <link href="/static/browse/0.3.4/css/latexml_styles.css" rel="stylesheet" type="text/css"/> <script src="https://cdn.jsdelivr.net/npm/bootstrap@5.3.0/dist/js/bootstrap.bundle.min.js"></script> <script src="https://cdnjs.cloudflare.com/ajax/libs/html2canvas/1.3.3/html2canvas.min.js"></script> <script src="/static/browse/0.3.4/js/addons_new.js"></script> <script src="/static/browse/0.3.4/js/feedbackOverlay.js"></script> <base href="/html/2507.07510v1/"/></head> <body> <nav class="ltx_page_navbar"> <nav class="ltx_TOC"> <ol class="ltx_toclist"> <li class="ltx_tocentry ltx_tocentry_section"><a class="ltx_ref" href="https://arxiv.org/html/2507.07510v1#S1" title="In Divergence Minimization Preference Optimization for Diffusion Model Alignment">1 Introduction</a></li> <li class="ltx_tocentry ltx_tocentry_section"> <a class="ltx_ref" href="https://arxiv.org/html/2507.07510v1#S2" title="In Divergence Minimization Preference Optimization for Diffusion Model Alignment">2 Background</a> <ol class="ltx_toclist ltx_toclist_section"> <li class="ltx_tocentry ltx_tocentry_subsection"><a class="ltx_ref" href="https://arxiv.org/html/2507.07510v1#S2.SS1" title="In 2 Background ‣ Divergence Minimization Preference Optimization for Diffusion Model Alignment">2.1 Diffusion Models</a></li> <li class="ltx_tocentry ltx_tocentry_subsection"><a class="ltx_ref" href="https://arxiv.org/html/2507.07510v1#S2.SS2" title="In 2 Background ‣ Divergence Minimization Preference Optimization for Diffusion Model Alignment">2.2 Reinforcement Learning from Human Feedback (RLHF)</a></li> </ol> </li> <li class="ltx_tocentry ltx_tocentry_section"> <a class="ltx_ref" href="https://arxiv.org/html/2507.07510v1#S3" title="In Divergence Minimization Preference Optimization for Diffusion Model Alignment">3 Method</a> <ol class="ltx_toclist ltx_toclist_section"> <li class="ltx_tocentry ltx_tocentry_subsection"><a class="ltx_ref" href="https://arxiv.org/html/2507.07510v1#S3.SS1" title="In 3 Method ‣ Divergence Minimization Preference Optimization for Diffusion Model Alignment">3.1 Revisiting Diffusion-DPO</a></li> <li class="ltx_tocentry ltx_tocentry_subsection"><a class="ltx_ref" href="https://arxiv.org/html/2507.07510v1#S3.SS2" title="In 3 Method ‣ Divergence Minimization Preference Optimization for Diffusion Model Alignment">3.2 Our Method</a></li> <li class="ltx_tocentry ltx_tocentry_subsection"><a class="ltx_ref" href="https://arxiv.org/html/2507.07510v1#S3.SS3" title="In 3 Method ‣ Divergence Minimization Preference Optimization for Diffusion Model Alignment">3.3 Theoretical Analysis</a></li> </ol> </li> <li class="ltx_tocentry ltx_tocentry_section"><a class="ltx_ref" href="https://arxiv.org/html/2507.07510v1#S4" title="In Divergence Minimization Preference Optimization for Diffusion Model Alignment">4 Related Work</a></li> <li class="ltx_tocentry ltx_tocentry_section"> <a class="ltx_ref" href="https://arxiv.org/html/2507.07510v1#S5" title="In Divergence Minimization Preference Optimization for Diffusion Model Alignment">5 Experiments</a> <ol class="ltx_toclist ltx_toclist_section"> <li class="ltx_tocentry ltx_tocentry_subsection"><a class="ltx_ref" href="https://arxiv.org/html/2507.07510v1#S5.SS1" title="In 5 Experiments ‣ Divergence Minimization Preference Optimization for Diffusion Model Alignment">5.1 Experimental Setting</a></li> <li class="ltx_tocentry ltx_tocentry_subsection"><a class="ltx_ref" href="https://arxiv.org/html/2507.07510v1#S5.SS2" title="In 5 Experiments ‣ Divergence Minimization Preference Optimization for Diffusion Model Alignment">5.2 Alignment Result</a></li> <li class="ltx_tocentry ltx_tocentry_subsection"><a class="ltx_ref" href="https://arxiv.org/html/2507.07510v1#S5.SS3" title="In 5 Experiments ‣ Divergence Minimization Preference Optimization for Diffusion Model Alignment">5.3 Image Editing Results</a></li> <li class="ltx_tocentry ltx_tocentry_subsection"><a class="ltx_ref" href="https://arxiv.org/html/2507.07510v1#S5.SS4" title="In 5 Experiments ‣ Divergence Minimization Preference Optimization for Diffusion Model Alignment">5.4 Ablation Study</a></li> </ol> </li> <li class="ltx_tocentry ltx_tocentry_section"><a class="ltx_ref" href="https://arxiv.org/html/2507.07510v1#S6" title="In Divergence Minimization Preference Optimization for Diffusion Model Alignment">6 Conclusion</a></li> <li class="ltx_tocentry ltx_tocentry_appendix"> <a class="ltx_ref" href="https://arxiv.org/html/2507.07510v1#A1" title="In Divergence Minimization Preference Optimization for Diffusion Model Alignment">A Proofs and Derivations</a> <ol class="ltx_toclist ltx_toclist_appendix"> <li class="ltx_tocentry ltx_tocentry_subsection"><a class="ltx_ref" href="https://arxiv.org/html/2507.07510v1#A1.SS1" title="In Appendix A Proofs and Derivations ‣ Divergence Minimization Preference Optimization for Diffusion Model Alignment">A.1 Proof of Theorem 1</a></li> <li class="ltx_tocentry ltx_tocentry_subsection"><a class="ltx_ref" href="https://arxiv.org/html/2507.07510v1#A1.SS2" title="In Appendix A Proofs and Derivations ‣ Divergence Minimization Preference Optimization for Diffusion Model Alignment">A.2 Convexity of Equation 12</a></li> <li class="ltx_tocentry ltx_tocentry_subsection"><a class="ltx_ref" href="https://arxiv.org/html/2507.07510v1#A1.SS3" title="In Appendix A Proofs and Derivations ‣ Divergence Minimization Preference Optimization for Diffusion Model Alignment">A.3 Proof of Theorem 2</a></li> </ol> </li> <li class="ltx_tocentry ltx_tocentry_appendix"> <a class="ltx_ref" href="https://arxiv.org/html/2507.07510v1#A2" title="In Divergence Minimization Preference Optimization for Diffusion Model Alignment">B Experimental Details</a> <ol class="ltx_toclist ltx_toclist_appendix"> <li class="ltx_tocentry ltx_tocentry_subsection"><a class="ltx_ref" href="https://arxiv.org/html/2507.07510v1#A2.SS1" title="In Appendix B Experimental Details ‣ Divergence Minimization Preference Optimization for Diffusion Model Alignment">B.1 Details of Evaluation</a></li> <li class="ltx_tocentry ltx_tocentry_subsection"><a class="ltx_ref" href="https://arxiv.org/html/2507.07510v1#A2.SS2" title="In Appendix B Experimental Details ‣ Divergence Minimization Preference Optimization for Diffusion Model Alignment">B.2 Details of Datasets</a></li> <li class="ltx_tocentry ltx_tocentry_subsection"><a class="ltx_ref" href="https://arxiv.org/html/2507.07510v1#A2.SS3" title="In Appendix B Experimental Details ‣ Divergence Minimization Preference Optimization for Diffusion Model Alignment">B.3 Pseudo code of DMPO</a></li> </ol> </li> <li class="ltx_tocentry ltx_tocentry_appendix"> <a class="ltx_ref" href="https://arxiv.org/html/2507.07510v1#A3" title="In Divergence Minimization Preference Optimization for Diffusion Model Alignment">C More Experimental Results</a> <ol class="ltx_toclist ltx_toclist_appendix"> <li class="ltx_tocentry ltx_tocentry_subsection"><a class="ltx_ref" href="https://arxiv.org/html/2507.07510v1#A3.SS1" title="In Appendix C More Experimental Results ‣ Divergence Minimization Preference Optimization for Diffusion Model Alignment">C.1 More Qualitative Results</a></li> <li class="ltx_tocentry ltx_tocentry_subsection"><a class="ltx_ref" href="https://arxiv.org/html/2507.07510v1#A3.SS2" title="In Appendix C More Experimental Results ‣ Divergence Minimization Preference Optimization for Diffusion Model Alignment">C.2 More Quantitative Results</a></li> </ol> </li> </ol></nav> </nav> <div class="ltx_page_main"> <div class="ltx_page_content"> <article class="ltx_document ltx_authors_1line">11footnotetext: Equal contribution; junior auth

完整 Prompt

Above I've shared:
(1) the full paper text,
(2) all paper figures labeled by figure number,
(3) the caption for the central poster figure I'm building.

TASK: This is a CONFERENCE POSTER. **NOT** an academic-paper figure.
Style requirements:

  - Multi-section layout with a clear poster structure: large title banner
    at the top with the paper title + author/affiliation strip, then 3-6
    distinct content panels arranged in columns or a grid.
  - Large legible fonts (text must be readable at 2 m viewing distance) —
    headings ≥ 60 pt visual size in the final image.
  - Use colour blocks / panel backgrounds to delineate sections (this is
    what makes it a poster, not a single-figure diagram).
  - Aspect ratio: portrait or landscape rectangle, NOT square.

If your output looks like a standard academic-paper figure (single panel,
no title banner, dense small text, no colour blocks), you've failed the
task. Render the COMPLETE poster, not just the central figure.

Just give me the final poster image.

立即试用此 Prompt

在生成器中自动预填此 prompt。

试用此 Prompt

DSPO Score-Matching Alignment for Diffusion Models

论文上下文

完整 Prompt

立即试用此 Prompt

相关 Prompt

PINN Framework Pipeline for Microscopy Data

DP-FTRL Privacy Amplification Analysis — Conference Poster

NC-TTT: Noise-Contrastive Test-Time Training Poster