syedtihaamahmad / atttacksonvideoobjectdetection

AtttacksonVideoObjectDetection

Test videos and attack video results are saved in assets folder in .mp4 formats

Setup python3.6 conda env, install requirements and run demo_ssd300.ipynb

Adversarial gradient attack based on SSD300

Benign , Misclassifying attack , BBOX Vanishing attack respectively

Setup python3.7 conda env, install requirements and run demo_patch.ipynb

GAN based attack based on YOLO

Training dataset : INRIA person dataset

link

Video Attack

Benign , Random-patched attack , GAN-patched attack respectively

Picture attack

Acknowlegement

This project is developed based on the following repositories and paper:

TOG

YOLO

paper