AI Paper Atlas

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

Towards Adversarially Robust Object Detection

Haichao Zhang, Jianyu Wang

ICCV 2019 · Adversarial robustness and security › Attacks on text and graphs · approach: Biologically inspired robust vision · 163 citations (OpenAlex)

Read the paper

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
adversarial training with task-oriented multi-task domain (MTD) instead of noneSSD VGG16 · VOC07 test, RAP attack eps 8mAP: 6.6 -> 44.9Improved
adversarial training with MTDSSD VGG16 · VOC07 test, DAG attack eps 8mAP: 0.3 -> 28.5Improved
adversarial training with MTDSSD VGG16 · VOC07 test, cleanmAP: 72.1 -> 48.0Hurt
use task-oriented domain MTD instead of task-agnostic domain CON (ablation)SSD VGG16 · VOC07 test, loss_cls PGD attackmAP: 18.3 -> 29.1Improved
use task-oriented domain MTD instead of task-agnostic domain CON (ablation)SSD VGG16 · VOC07 test, loss_loc PGD attackmAP: 27.2 -> 31.9Improved
use task-oriented domain MTD instead of task-agnostic domain CON (ablation)SSD VGG16 · VOC07 test, cleanmAP: 38.7 -> 48.0Improved
use multi-task domain MTD instead of classification-only domain CLS (ablation)SSD VGG16 · VOC07 test, loss_cls PGD attackmAP: 21.8 -> 29.1Improved
use multi-task domain MTD instead of localization-only domain LOC (ablation)SSD VGG16 · VOC07 test, DAG attackmAP: 17.2 -> 28.5Improved
use multi-task domain MTD instead of CLS (ablation)SSD VGG16 · VOC07 test, loss_loc PGD attackmAP: 32.2 -> 31.9No clear effect
use CLS domain instead of LOC domain (ablation)SSD VGG16 · VOC07 test, cleanmAP: 51.9 vs 46.7Mixed
adversarial training with MTD on ResNet50 backboneSSD ResNet50 · VOC07 test, DAG attackmAP: 0.4 -> 22.9Improved
adversarial training with MTD on DarkNet53 backboneSSD DarkNet53 · VOC07 test, RAP attackmAP: 8.2 -> 46.6Improved

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