winterwindwang / Full-coverage-camouflage-adversarial-attack

https://winterwindwang.github.io/Full-coverage-camouflage-adversarial-attack/

winterwindwang/Full-coverage-camouflage-adversarial-attack Issues

how to extract images from train folder？
Updated 7 months ago1
camounflaged car always be detected in one angle.
Updated 7 months ago1
The resulting texture is very shallow
Closed 2 years ago10
在运行神经渲染器的save_obj中，遇到cuda问题
Closed 2 years ago4
有关相机位置和角度的问题
Closed a year ago5
渲染过后的图片小于mask的大小
Closed a year ago10
您提供的train_label_new标注好像有问题，x_cener, y_center完全相同
Updated 7 months ago2
您好，我想试着跑一下这个程序，但readme上没有具体的步骤，可以麻烦您提供一下嘛，万分感谢！
Updated 9 months ago
请问贵实验室是否还招收博士
Updated 9 months ago
想问一下，大佬用的训练模型的权重是自己用coco数据集预训练得到的吗？如果不是的话使用的是yolov3哪个版本的release权重呢
Closed 9 months ago
作者您好，我看到您的项目assets中有实物打印的模型，我想请问一下您模型在哪里购买的呢？
Closed a year ago
作者您好，我想问一下，您的训练最后的总损失大概在多少浮动
Closed a year ago
FCA训练只可以单GPU训练吗？多GPU运行到对dataloader进行迭代时会出现CUDA error: initialization error，这部分有一定的限制不能够多GPU吗？
Closed a year ago2
How can I get the obj file after attaching the texture
Closed 2 years ago2
我想请问下fca中用来测试的yolov5模型，faster-rcnn模型以及SSD模型他们的置信度阈值都是0.5嘛
Updated a year ago1
attack performance
Updated 2 years ago1
大佬我想问下生成攻击纹理的代码是哪个呀？孩子实在有些看懵了....
Updated 2 years ago1
total_img errors occur like this
Closed 2 years ago2
pytorch
Closed 2 years ago16
The ground truth?
Closed 2 years ago3
DAS and FCA training consult
Closed 2 years ago3
I would like to ask the author, is neuralRender installed in the same way as DAS? Thank you!
Closed 2 years ago2
train_new
Closed 2 years ago3
DAS
Closed 2 years ago2
npy文件生成
Closed 2 years ago6
渲染过程细节
Closed 2 years ago10
Do chainer and other packages have version installation requirements? or it's better to follow das？
Closed 2 years ago2
I would like to ask the author whether the training interface is consistent with your training at that time. Because the loss is still oscillating in the later rounds, and the grad mean of the texture has always been 0, I am a little worried about my training
Closed 2 years ago4
texture转图片
Closed 2 years ago
May I ask if the author has successfully printed the texture picture? Thank you for your reply！
Closed 2 years ago1
Are the first three quantities of cam_trans in npz relative to the coordinates of the car ？
Closed 2 years ago
May I ask whether the detection model used for the test in the paper is Yolov5L.pt
Closed 2 years ago
1
Closed 2 years ago
The loss during the training increased. After the training of 10 epochs, the loss was about 1.1. Is this normal?Thanks for answering.
Closed 2 years ago8
TypeError: __init__() got an unexpected keyword argument 'camera_mode'
Closed 2 years ago
TypeError: transpose() received an invalid combination of arguments - got (tuple), but expected one of: * (name dim0, name dim1) * (int dim0, int dim1)
Closed 2 years ago
TypeError: transpose() received an invalid combination of arguments - got (tuple), but expected one of:
Closed 2 years ago1
a
Closed 2 years ago
KeyError: 'msgs'
Closed 2 years ago1
AttributeError: 'numpy.ndarray' object has no attribute 'device'
Closed 2 years ago
FileNotFoundError: [Errno 2] No such file or directory: 'data/phy_attack/train/data10067.npz'，但指定的npz路径(das项目下)有这个文件来着
Closed 2 years ago