新的隐私防御方法用于修剪视觉语言模型的视觉标记

作者 PulseAugur 编辑部 · [1 个来源] · 2026-09-15 04:00

研究人员开发了 QPriv-VL，一个旨在增强用于联邦学习等敏感应用的视觉语言模型（VLM）隐私的新框架。该系统在传输前智能地修剪视觉标记，从而降低数据传输成本并减少暴露私人信息的风险。QPriv-VL 使用动态阈值预测器（DTP），该预测器考虑问题相关性和特征敏感性，以选择性地保留重要的视觉数据，同时抑制潜在的敏感区域，在传输的标记少得多的情况下实现了具有竞争力的准确性。 AI

影响增强了视觉语言模型在敏感应用中的隐私和效率。

排序理由详细介绍视觉语言模型新方法的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.AI 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 15 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 详细介绍视觉语言模型新方法的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: paper, safety; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Breaking (< 6h); Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv cs.AI TIER_1 English(EN) · Md Khalid Syfullah, Alvi Ataur Khalil · 2026-09-15 04:00

无需发送不必要内容：问题引导的令牌修剪作为视觉语言模型的隐私防御

arXiv:2609.15671v1 Announce Type: cross Abstract: Visual Question Answering (VQA) with Vision-Language Models (VLMs) is increasingly used in privacy-sensitive and bandwidth-constrained settings. Federated Learning (FL), Split Learning (SL), and U-Shaped Split Learning (USL) keep …

本文如何被排名

报道来源 [1]

无需发送不必要内容：问题引导的令牌修剪作为视觉语言模型的隐私防御

相关实体

相关话题