A global adversarial example generation method based on class gradients in a single picture and, in a wider scope, a proportion of distributed data in the dataset. The designed global perturbation can affect all data in the dataset. The performance of the proposed method in the PascalVOC and MS-COCO-keypoints datasets is better than that of the existing methods.

Overview

Global_-Category-wise_Attack

Getting Started

This document provides tutorials of Global Catergory-wise Attack (GCA). Basically it contains two parts:

  1. The installation procedures of the target network CenterNet, more details can be found at Object as Points
  2. The installation procedures of GCA.
  3. Make sure you have finished Installation Procedures and Datasets before started.

Networks download

Download the models you want to evaluate from the model zoo and put them in Root_File/models/.

##Evaluation

Detection

To evaluate PascalVOC object detection

> Clean outputs:
> run:
python test.py ctdet --exp_id pascal_dla_1x_clean --dataset pascal --arch dla_34 --not_prefetch_test --load_model ../models/ctdet_pascal_dla_512.pth
> Adversarial outputs (GCA):
> run:
python test_universal.py ctdet --exp_id pascal_dla_1x_Uniattack --dataset pascal --arch dla_34 --not_prefetch_test --load_model ../models/ctdet_pascal_dla_512.pth

Pose estimation

> Clean outputs:
> run:
python test.py multi_pose --exp_id coco_resdcn18_clean --dataset coco_hp --arch dla_34 --not_prefetch_test --load_model ../models/multi_pose_dla_1x.pth
> Adversarial outputs (GCA):
> run:
python test_universal.py multi_pose --exp_id coco_resdcn18_attack_uni --dataset coco_hp --arch dla_34 --not_prefetch_test --load_model ../models/multi_pose_dla_1x.pth
You might also like...

Make an example. Demonstrate performance speed-up. Demonstrate scalability - how does performance changes with the number of cores?

This example is part of an assignment for the course Concurrent Computing in Robotics from JEMARO. OpenCL - First Steps Head to the report document fo

Mar 11, 2022

A proposed standard `NOCK` for a Parquet format that supports efficient distributed serialization of multiple kinds of graph technologies

pynock The following describes a proposed standard NOCK for a Parquet format that supports efficient distributed serialization of multiple kinds of gr

Dec 1, 2022

KeypointNeRF Generalizing Image-based Volumetric Avatars using Relative Spatial Encoding of Keypoints

KeypointNeRF Generalizing Image-based Volumetric Avatars using Relative Spatial Encoding of Keypoints

KeypointNeRF: Generalizing Image-based Volumetric Avatars using Relative Spatial Encoding of Keypoints Marko Mihajlovic · Aayush Bansal · Michael Zoll

Dec 27, 2022

Bamboo: 4 times larger than ImageNet; 2 time larger than Object365; Built by active learning.

Bamboo: 4 times larger than ImageNet; 2 time larger than Object365; Built by active learning.

Zhang Yuanhan1  Sun Qinghong2  Zhou Yichun3  He Zexin3  Yin Zhenfei4  Wang Kun4  Sheng Lv3  Qiao Yu5  Shao Jing4  Liu Ziwei1 1S-Lab, Nanyang Technolog

Dec 24, 2022

Designed a containerized pipeline to detect fake-face submissions using a CNN model including generation of fake datasets for machine learning in a Test Driven Development environment.

Designed a containerized pipeline to detect fake-face submissions using a CNN model including generation of fake datasets for machine learning in a Test Driven Development environment.

Sep 12, 2022

Code and data for the paper *Better Smatch = Better Parser? AMR evaluation is not so simple anymore*

AMR parse(r) evaluation experiment This is code/data for our paper Better Smatch = better parser? AMR evaluation is not so simple anymore Data Data fo

Dec 21, 2022

Code for T-Few from "Few-Shot Parameter-Efficient Fine-Tuning is Better and Cheaper than In-Context Learning"

T-Few This repository contains the official code for the paper: "Few-Shot Parameter-Efficient Fine-Tuning Outperforms In-Context Learning". This metho

Dec 31, 2022

⚡A python script that detects objects in images using YOLOv3 trained on the coco dataset

⚡A python script that detects objects in images using YOLOv3 trained on the coco dataset

This is a Python Script which uses YoloV3 to detect objects in images. YOLOv3 (You Only Look Once, Version 3) is a real-time object detection algorith

Apr 13, 2022

Implementation of RQ Transformer, proposed in the paper "Autoregressive Image Generation using Residual Quantization"

Implementation of RQ Transformer, proposed in the paper

RQ-Transformer Implementation of RQ Transformer, which proposes a more efficient way of training multi-dimensional sequences autoregressively. This re

Dec 6, 2022
Owner
Yancy
Yancy
A pure Tensorflow2.0 implementation of EfficientDet. Run on PascalVOC or your own datasets too!

Tensorflow2.0 EfficientDet A pure WORKING Tensorflow2.0 implementation of EfficientDet for object detection. There are too many non-working versions o

Thomas Chia 4 Jul 2, 2022
Integrate mypy in seconds with existing codebase. A friendly CLI tool to make mypy report only new type violations and ignore existing ones.

mypy-baseline A CLI tool for painless integration of mypy with an existing Python project. When you run it for the first time, it will remember all ty

Orsinium Labs 17 Dec 28, 2022
PyTorch Implementation for Robust Perturbation for Visual Explanation: Cross-Checking Mask Optimization to Avoid Class Distortion

IEEE Transactions on Image Processing Robust Perturbation for Visual Explanation: Cross-Checking Mask Optimization to Avoid Class Distortion Authors:

Junho Kim 2 May 25, 2022
Adansons Base provides a better way to management unstructured data. It helps to organize metadata and create datasets to improve AI performance.

Adansons Base Document Product Concept 1. Installation 2. Configuration 2.1 with CLI 2.2 Environment Variables 3. Tutorial 1: Organize meta data and C

Adansons Inc 27 Oct 22, 2022
I proposed a new way, Momentum Prototype UDA to solve Unsupervised Domain Adaptation problem on BraTS2021 dataset. It achieves a similar performance when compared to SOTA solutions.

Momentum-Prototype-Unsupervised-Domain-Adaptation Why Unsupervised Domain Adaptation The traditional deep learning relies on a lot of labelled data to

Wenqing Zong 1 Oct 12, 2022
An easy to use roblox cookie logger. This cookie logger roblox logs better than all other cookie logger roblox!

CookieLoggerRoblox This cookie logger for roblox is the best out there. This roblox cookie logger uses a stub in order to log people! What is this bot

1992honda 1 Aug 22, 2022
Python script used to investigate which Indonesian online news outlet is most likely to affect public opinion in Indonesia.

NewMediaChurn Research Project This repository contains a python script that collects Indonesian headline news on 10th of August 2022 using the newsap

Keanu Nurherbyanto 1 Aug 10, 2022
Implementation of AAAI 2022 Paper: Go wider instead of deeper

WideNet_Code Implementation of AAAI 2022 Paper: Go wider instead of deeper arXiv You can run our code in this way. $ cd WideNet_Code $ bash submit_tpu

Xue Fuzhao 5 Dec 25, 2022
API for the dataset proposed in "Pose2UV: Single-shot Multi-person Mesh Recovery with Deep UV Prior" @ TIP2022.

3DMPB-dataset Dataset proposed in "Pose2UV: Single-shot Multi-person Mesh Recovery with Deep UV Prior" [ProjectPage] [paper] @ TIP2022. Samples of 3DM

null 20 Jul 27, 2022