Skip to content

Translation in progress Some entries still include text in the original language.

Oct 12 · Monday

All papers

Found 1 paper
Filter3

Attack type

Poisoning and backdoors10Extraction and theft3Detection evasion3Malicious use1

Layer

Prompt layer4Model layer5Training and data layer1

System component

Multimodal1
Select any value within a category; different categories must all match. Counts include the other filters you have selected.Another 247papers have no tags yet and are not included in filtered results.
Attack type：Malicious useLayer：Training and data layerSystem component：MultimodalClear all

Another 247papers do not have category tags yet and are not included in filtered results.

  1. DefensesarXiv 2610.03980

    FADE：面向视频扩散模型的帧感知多概念擦除框架

    提出 FADE，对文生视频模型做帧感知多概念擦除

    arXiv
    2610.03980
    Published
    Layer
    Training and data layer
    Setting
    Models and APIs
    Evaluated models
    Wan2.1-T2V-1.3B, Wan2.1-T2V-14B, CogVideoX-2B and others4
    AbstractFADE 用闭式编辑、帧感知适配器和软路由，在单骨干上擦除多概念并压低逐帧残差。

    FADE 是面向文生视频扩散 Transformer 的多概念遗忘方法，用来抑制指定物体、画风、裸体、暴力、仇恨和名人身份，并尽量保持其他生成。作者要处理两件事。与帧无关的抑制会让概念在少数帧重现，片段均值会掩盖这一失败。

    Paper analysisRead analysis
You have reached the end