StoryTailor: 一种零样本 pipelines 用于动作丰富的多主体视觉叙事
StoryTailor:A Zero-Shot Pipeline for Action-Rich Multi-Subject Visual Narratives
AI总结：
StoryTailor通过零样本方法生成动作丰富的多主体视觉叙事，结合GCA、AB-SVR和SFC模块提升动作忠实性和跨帧连续性，实验显示其在多个指标上优于基线方法。
摘要：
生成无需微调的多帧、动作丰富的视觉叙事面临三重矛盾：动作文本的忠实性、主体身份的保真度以及跨帧背景的连续性。我们提出了StoryTailor，一个在单块RTX 4090（24GB）上运行的零样本流水线，能够从长叙事提示、每主体参考和定位框生成时间上连贯、身份保持的图像序列。三个协同模块驱动该系统：Gaussian-Centered Attention（GCA）动态聚焦每个主体核心并缓解定位框重叠；Action-Boost Singular Value Reweighting（AB-SVR）在文本嵌入空间中放大与动作相关的方向；以及Selective Forgetting Cache（SFC）保留可转移的背景线索，遗忘非必要的历史，并选择性地呈现保留的线索以构建跨场景语义联系。与基线方法相比，实验显示CLIP-T在10-15%范围内有所提升，DreamSim低于强基线，而CLIP-I在视觉可接受且竞争范围内。在24GB GPU上匹配分辨率和步骤的情况下，推理速度比FluxKontext更快。在定性方面，StoryTailor能够提供富有表现力的交互和演进但稳定的场景。
英文摘要：
Generating multi-frame, action-rich visual narratives without fine-tuning faces a threefold tension: action text faithfulness, subject identity fidelity, and cross-frame background continuity. We propose StoryTailor, a zero-shot pipeline that runs on a single RTX 4090 (24 GB) and produces temporally coherent, identity-preserving image sequences from a long narrative prompt, per-subject references, and grounding boxes. Three synergistic modules drive the system: Gaussian-Centered Attention (GCA) to dynamically focus on each subject core and ease grounding-box overlaps; Action-Boost Singular Value Reweighting (AB-SVR) to amplify action-related directions in the text embedding space; and Selective Forgetting Cache (SFC) that retains transferable background cues, forgets nonessential history, and selectively surfaces retained cues to build cross-scene semantic ties. Compared with baseline methods, experiments show that CLIP-T improves by up to 10-15%, with DreamSim lower than strong baselines, while CLIP-I stays in a visually acceptable, competitive range. With matched resolution and steps on a 24 GB GPU, inference is faster than FluxKontext. Qualitatively, StoryTailor delivers expressive interactions and evolving yet stable scenes.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
