FaceBoxes.PyTorch
FaceBoxes.PyTorch copied to clipboard
zisianw
A PyTorch Implementation of FaceBoxes
我参与的项目中需要每次同时处理两张图像，于是我将两张图像拼接为一张输入到faceboxes中，但这种方式的时间开销比分别处理两张图像所需的时间更大。有没有更加推荐的方式呢？
I have reimplement FaceBoxes with c++, [faceboxes.cpp](https://github.com/DefTruth/lite.ai/blob/main/lite/ort/cv/faceboxes.cpp). It seems FaceBoxes suitable for mobile device with some simple frontal faces, but can not get good accuracy for complicated situation.
I am currently trying to reduce the number of prior_boxes. The current approach is to reduce the dense anchor ratio to 2:1:1 (originally 4:2:1). This scheme is currently available on...
我是用来检测球，因为用原始的，需要放大2.5倍，才能用来检测乒乓球，才检测得到，这时帧率变得很低，所以我想调小这个 cfg = { 'name': 'FaceBoxes', #'min_dim': 1024, #'feature_maps': [[32, 32], [16, 16], [8, 8]], # 'aspect_ratios': [[1], [1], [1]], 'min_sizes': [[32, 64, 128], [256], [512]], # 'min_sizes': [[8, 16,...
Can i use the same code on PASCAL VOC dataset where num_classes = 20, by changing the same parameter in your code, Will it work? since you have used RPN...
@zisianw @sfzhang15 为了表达清楚，这里就用中文了！ 关于train loss中的震荡现象． 自己基于提供的代码修改参数进行了3次训练，3次的差异在于batch_size和LR．每次训练时总的iterations保持一样（120K）． 唯一的差异是自己基于单GPU进行 在默认参数下，即：batch_size=32，learning rate = 0.01，epoches = 300．训练．最终结果：PASCAL：0.9630；AFW：0.9838；FDDB：0.954/0.725； 训练的cls loss和reg loss分别为下图： [trian_loss_C_FaceBoxes1.pdf](https://github.com/zisianw/FaceBoxes.PyTorch/files/4809262/trian_loss_C_FaceBoxes1.pdf) [trian_loss_L_FaceBoxes1.pdf](https://github.com/zisianw/FaceBoxes.PyTorch/files/4809263/trian_loss_L_FaceBoxes1.pdf) 修改参数，即： batch_size=16，learning rate = 0.005，epoches = 150．训练．最终结果：PASCAL：0.9628；AFW：0.9826；DDB：0.954/0.724． 得到的cls loss和reg loss分别为下图： [trian_loss_C_FaceBoxes4.pdf](https://github.com/zisianw/FaceBoxes.PyTorch/files/4809265/trian_loss_C_FaceBoxes4.pdf) [trian_loss_L_FaceBoxes4.pdf](https://github.com/zisianw/FaceBoxes.PyTorch/files/4809266/trian_loss_L_FaceBoxes4.pdf)...
