AI Paper Atlas

Research map / Adversarial robustness and security / Attacks on text and graphs

TrojVLM: Backdoor Attack Against Vision Language Models

Weimin Lyu, Lu Pang, Tengfei Ma, Haibin Ling, Chao Chen

ECCV 2024 · Adversarial robustness and security › Attacks on text and graphs · approach: Vision-language model robustness · 25 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add semantic preservation loss to language modeling loss (ablation)TrojVLM · Flickr8k captioning (poisoned images, sentence target)CIDEr: 108.4 -> 114.3Improved
add semantic preservation loss to language modeling loss (ablation)TrojVLM · Flickr8k captioning (poisoned images, word target)ASR: 0.961 -> 0.976Improved
add semantic preservation loss to language modeling loss (ablation)TrojVLM · OK-VQA (poisoned images, word target)VQA score: 42.7 -> 43.7Improved
use TrojVLM instead of Blended baseline for text qualityTrojVLM · Flickr8k captioning (poisoned images)CIDEr: 4.5 -> 114.3Improved
use TrojVLM instead of AnyDoor on poisoned imagesTrojVLM · OK-VQA (poisoned images)VQA score: 41.9 -> 45.7Improved
BadEncoder (CLIP encoder attack) applied to VLM captioningBadEncoder · Flickr8k captioning (poisoned images)ASR: 0.000Hurt
increase trigger size from 5 to 30 pixels (ablation)TrojVLM · Flickr8k captioningASR: 0.664 -> 1.000Improved
increase poison rate from 0.05 to 0.1 (ablation)TrojVLM · Flickr8k captioningCIDEr: 109.1 -> 114.3Improved
increase poison rate from 0.1 to 0.3 (ablation)TrojVLM · Flickr8k captioningCIDEr: 114.3 -> 114.7No clear effect
change trigger location from upper left to other positions (ablation)TrojVLM · Flickr8k captioningASR: 0.979 -> 0.999No clear effect
keep only the 1% of image tokens containing the trigger (ablation)TrojVLM · Flickr8k captioningASR: 0.96 (word), 0.97 (sent), 0.90 (web)No clear effect
apply TrojVLM to MiniGPT-4 and InstructBLIPTrojVLM · Flickr8k captioningASR: 0.959 to 0.986Improved

Later papers that compared with this method

1 later accepted paper reports TrojVLM as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy