{"id":13644975,"url":"https://github.com/VITA-Group/3D_Adversarial_Logo","last_synced_at":"2025-04-21T11:32:16.782Z","repository":{"id":86240792,"uuid":"273955778","full_name":"VITA-Group/3D_Adversarial_Logo","owner":"VITA-Group","description":"[Preprint] \"Can 3D Adversarial Logos Cloak Humans?\" ","archived":false,"fork":false,"pushed_at":"2021-12-30T12:41:46.000Z","size":3490,"stargazers_count":18,"open_issues_count":1,"forks_count":3,"subscribers_count":6,"default_branch":"master","last_synced_at":"2025-04-19T18:51:30.836Z","etag":null,"topics":["3d-mesh","adversarial-attacks","cloak","detection","detector","logos"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/VITA-Group.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null}},"created_at":"2020-06-21T17:50:06.000Z","updated_at":"2024-10-16T03:43:17.000Z","dependencies_parsed_at":"2023-03-13T09:13:33.079Z","dependency_job_id":null,"html_url":"https://github.com/VITA-Group/3D_Adversarial_Logo","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/VITA-Group%2F3D_Adversarial_Logo","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/VITA-Group%2F3D_Adversarial_Logo/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/VITA-Group%2F3D_Adversarial_Logo/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/VITA-Group%2F3D_Adversarial_Logo/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/VITA-Group","download_url":"https://codeload.github.com/VITA-Group/3D_Adversarial_Logo/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":250047981,"owners_count":21366152,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["3d-mesh","adversarial-attacks","cloak","detection","detector","logos"],"created_at":"2024-08-02T01:02:22.600Z","updated_at":"2025-04-21T11:32:16.765Z","avatar_url":"https://github.com/VITA-Group.png","language":"Python","funding_links":[],"categories":["Object Detection Applications"],"sub_categories":[],"readme":"\n# Can 3D Adversarial Logos Clock Humans? #\n\n[![License: MIT](https://img.shields.io/badge/License-MIT-green.svg)](https://opensource.org/licenses/MIT)\n\n[Can 3D Adversarial Logos Clock Humans?]()\n\nYi Wang\\*, Jingyang Zhou*, Tianlong Chen, Sijia Liu, Shiyu Chang, Chandrajit Bajaj, Zhangyang Wang\n\n\n\n## Overview\n\nExamples of our 3D adversarial logo attack on different3D object meshes to fool a YOLOV2 detector. \n\n![](./doc_imgs/intro.png)\n\n\n\n## Methodology\n\n![](./doc_imgs/methods.png)\n\n\n\n## Prerequisites and Installation\n\n- Python 3\n\n- Pytorch 1.1.0\n\n- CUDA 9.2 (lower versions may work but were not tested)\n\n- TensorboardX \n\n  ```shell\n  pip install tensorboardX tensorboard\n  ```\n\n- Neural renderer \n\n  ```shell\n  pip install neural-renderer-pytorch\n  ```\n\n- Clone this repo \n\n  ```shell\n  git clone https://github.com/TAMU-VITA/3D_Adversarial_Logo.git\n  cd 3D_Adversarial_Logo\n  ```\n\n- Yolov2 weights\n\n  ```shell\n  mkdir weights\n  curl https://pjreddie.com/media/files/yolov2.weights -o weights/yolo.weights\n  ```\n\n  \n\n## Usage\n\n### Prepare the dataset\n\n- Datasets and our prediction images can be found [here](https://drive.google.com/file/d/1CUC_Yy10sjGSvSG8wt-ToPtOJ_eEf8x-/view?usp=sharing).\n- Our pre-trained 3d adversarial logos can be found [here](https://drive.google.com/file/d/1AeWOzhLUbf7XK4-bo-rtvcQ1VKO_bn30/view?usp=sharing).\n\nRemark. After downloading them, put the data and weight under the `3D_Adversarial_Logo` folder.\n\n\n\n### Arguments\n\n- `--width` : the width of the universal logo.\n- `--height` : the height of the universal logo.\n- `--depth` : the depth of the universal logo (default is 1 to form an image).\n- `--angle_range` : the angle range of the camera view for training (eg. 10 for +/- 10 degree)\n- `--logonum` : the total faces of the 3D logo.\n- `--train_mesh` : which mesh to train (the 123 denotes mesh 1\u00262\u00263, which can be further upgraded by adding `nargs='+'` in parsing args).\n- `--test_mesh` : which mesh to test (the 123 denotes mesh 1\u00262\u00263, which can be further upgraded by adding `nargs='+'` in parsing args).\n- `--consistent` : whether to share the same universal logo among all the meshes.\n- `--logo_ref` : which logo to train with.\n- `--train_patch` : whether to train 2D patch.\n- `--restore_model` : whether to restore the saved universal logo.\n- `--save_model` : whether to save the universal logo.\n\n\n\n### Run 3D adversarial logos ###\n\n- logo: X\n\n  ```shell\n  python train_patch1.py --width=100 --height=150 --depth=1 --angle_range=0 --logonum=25 --train_mesh=123 --consistent --logo_ref=X --save_model \u003elogo-records/logoX.out 2\u003e\u00261 \u0026\n  ```\n\n- logo: X (medium) \n\n  ```shell\n  python train_patch1.py --width=100 --height=150 --depth=1 --angle_range=0 --logonum=10 --train_mesh=123 --consistent --logo_ref=X --save_model \u003elogo-records/logoX.out 2\u003e\u00261 \u0026\n  ```\n\n- logo: X (small) \n\n  ```shell\n  python train_patch1.py --width=100 --height=150 --depth=1 --angle_range=0 --logonum=5 --train_mesh=123 --consistent --logo_ref=X --save_model \u003elogo-records/logoX.out 2\u003e\u00261 \u0026\n  ```\n\n- logo: H\n\n  ```shell\n  python train_patch1.py --width=100 --height=150 --depth=1 --angle_range=0 --logonum=25 --train_mesh=123 --consistent --logo_ref=H --save_model \u003elogo-records/logoH.out 2\u003e\u00261 \u0026\n  ```\n\n- logo: T\n\n  ```shell\n  python train_patch1.py --width=100 --height=150 --depth=1 --angle_range=0 --logonum=25 --train_mesh=123 --consistent --logo_ref=T --save_model \u003elogo-records/logoT.out 2\u003e\u00261 \u0026\n  ```\n\n- logo: G\n\n  ```shell\n  python train_patch1.py --width=100 --height=100 --depth=1 --angle_range=0 --logonum=17 --train_mesh=123 --consistent --logo_ref=G --save_model \u003elogo-records/logoG.out 2\u003e\u00261 \u0026\n  ```\n\n- logo: C \n\n  ```shell\n  python train_patch1.py --width=100 --height=100 --depth=1 --angle_range=0 --logonum=17 --train_mesh=123 --consistent --logo_ref=C --save_model \u003elogo-records/logoC.out 2\u003e\u00261 \u0026\n  ```\n\n- logo: O \n\n  ```shell\n  python train_patch1.py --width=100 --height=100 --depth=1 --angle_range=0 --logonum=17 --train_mesh=123 --consistent --logo_ref=O --save_model \u003elogo-records/logoO.out 2\u003e\u00261 \u0026\n  ```\n\n- logo: tw \n\n  ```shell\n  python train_patch1.py --width=100 --height=100 --depth=1 --angle_range=0 --logonum=25 --train_mesh=123 --consistent --logo_ref=tw --save_model \u003elogo-records/logotw.out 2\u003e\u00261 \u0026\n  ```\n\n- logo: drop \n\n  ```shell\n  python train_patch1.py --width=100 --height=150 --depth=1 --angle_range=0 --logonum=25 --train_mesh=123 --consistent --logo_ref=drop --save_model \u003elogo-records/logodrop.out 2\u003e\u00261 \u0026\n  ```\n\n  \n\n### Train 2D adversarial patch\n- logo: G \n\n  ```shell\n  python train_patch1.py --width=100 --height=100 --depth=1 --angle_range=0 --logonum=17 --train_mesh=123 --consistent --logo_ref=G --restore_model --save_model --train_patch\n  ```\n\n\n\n## Citation\n\nIf you are use this code for you research, please cite our paper.\n\n```\n@article{chen2020can,\n  title={Can 3D Adversarial Logos Cloak Humans?},\n  author={Chen, Tianlong and Wang, Yi and Zhou, Jingyang and Liu, Sijia and Chang, Shiyu and Bajaj, Chandrajit and Wang, Zhangyang},\n  journal={arXiv preprint arXiv:2006.14655},\n  year={2020}\n}\n```\n\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2FVITA-Group%2F3D_Adversarial_Logo","html_url":"https://awesome.ecosyste.ms/projects/github.com%2FVITA-Group%2F3D_Adversarial_Logo","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2FVITA-Group%2F3D_Adversarial_Logo/lists"}