Giter Site home page Giter Site logo

bmvc2018r-ap's Introduction

DenseFoolbox

DenseFoolbox is a python repository for white-box attacking object detectors, instance segmentation. This repository is a simple implementation of our paper Robust Adversarial Perturbation on Deep Proposal-based Models, BMVC2018.

Content

  1. Overview
  2. Requirements
  3. Demo

Overview

overview

We target Region Proposal Network (RPN) as the bottleneck of Deep-proposal based networks. The detections can be disrupted by breaking object proposal generation. To do so, we disturb the predicted class score as well as offset regression of object proposals.

Requirements

  • Pytorch 0.4.0
  • Ubuntu 16.04
  • CUDA 8.0
  • Python 2.7
  • opencv3

Demo

Attacking Faster-RCNN

  1. We use Faster-RCNN detector based on pytorch framework pytorch-faster-rcnn. We make modifications to this repository which can be downloaded herehere[code:rqic].
  2. Unzip the repository to object_detectors.
  3. Look into attack_wrapper/object_detectors_v2 and run run.py.
    python run.py \
    --net=faster-rcnn \  # faster-rcnn or ssd (update later)
    --base=vgg16 \
    --data_dir=demo/ \
    --res_dir=res/
    

Attacking Mask-RCNN

Update later

Citation

If you find this implementation helpful, please cite:

@inproceedings{li2018rap,
author={Li, Yuezun and Tian, Daniel and Chang, Mingching and Bian, Xiao and Lyu, Siwei},
title={Robust Adversarial Perturbation on Deep Proposal-based Models},
booktitle={BMVC},
year={2018}}    

bmvc2018r-ap's People

Contributors

yuezunli avatar

Stargazers

 avatar Jialin Lu avatar Jack avatar bobzhou avatar upbeatcorgi avatar c1assy avatar  avatar  avatar  avatar  avatar  avatar Chunxu Guo avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar Dylan avatar  avatar  avatar gggmt avatar

Watchers

 avatar

bmvc2018r-ap's Issues

cannot download modifications of pytorch-faster-rcnn repository

@DanMohaha Hi, I'm interested in your repository, but cannot download modifications of pytorch-faster-rcnn. I use wget and got this error:

Resolving doc-0s-10-docs.googleusercontent.com (doc-0s-10-docs.googleusercontent.com)... 216.58.200.225, 2404:6800:4008:800::2001
Connecting to doc-0s-10-docs.googleusercontent.com (doc-0s-10-docs.googleusercontent.com)|216.58.200.225|:443... failed: Connection timed out.
Connecting to doc-0s-10-docs.googleusercontent.com (doc-0s-10-docs.googleusercontent.com)|2404:6800:4008:800::2001|:443... failed: Network is unreachable

Downloading with Chrome is also failed. Can you please figure out this problem? I'm waiting for your reply, thanks.

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.