无需训练即可提升文本生成图像多样性，避免重复输出。
TPSO: Training-Free Diverse Image Generation via Semantic Prompt Embedding Optimization
- 通过优化提示词嵌入空间，在不训练模型的前提下探索低频生成区域。
- 在MS-COCO数据集上多样性评分从1.10提升至4.18，图像质量无明显下降。
- 适用于任意扩散模型，推理开销仅增加3.6%~8.9%，适合创意生成场景。
图像多样性仍是文生图扩散模型的核心挑战。低多样性生成常导致输出重复，增加采样冗余，影响创意探索与下游应用。其关键原因是扩散模型倾向于坍缩到学习分布中的强模式。现有提升多样性的方法如基于引导的策略，常引入失真，降低图像质量。为此，我们提出无需训练、模型无关的词元-提示嵌入空间优化（TPSO）模块。TPSO引入可学习参数，探索词元嵌入空间中未充分覆盖的区域，减少对强模式的重复采样。同时，提示级语义约束调控分布偏移，防止质量下降并保持语义一致性。在三个代表性扩散模型上对MS-COCO数据集的大量实验表明，TPSO显著提升多样性，评分从1.10增至4.18，同时保持图像质量，推理时间仅增加3.6%至8.9%。代码已公开于：https://github.com/Open-Debin/TPSO。
原文摘要 · Abstract (English)
Image diversity remains a fundamental challenge for text-to-image diffusion models. Low-diversity generation often leads to repetitive outputs, increasing sampling redundancy and hindering both creative exploration and downstream applications. A key factor is the tendency of diffusion models to collapse toward strong modes in the learned distribution. Existing attempts to improve diversity, such as steering-based guidance, often introduce distortions that degrade image quality. To address this issue, we propose Token-Prompt Embedding Space Optimization (TPSO), a training-free and model-agnostic module. TPSO introduces learnable parameters to explore underrepresented regions of the token embedding space, reducing the tendency to repeatedly sample from strong modes of the distribution. Meanwhile, a prompt-level semantic constraint regulates distribution shifts, preventing quality degradation while preserving semantic fidelity. Extensive experiments on MS-COCO across three representative diffusion backbones demonstrate that TPSO substantially improves diversity, boosting performance from 1.10 to 4.18, while maintaining image quality with only a modest inference-time overhead of 3.6% to 8.9%. Code is available at: https://github.com/Open-Debin/TPSO.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。