yja1

load_embeddings.py run error

1

d = len(all_vocab['the']) when run this line occur error, because all_vocab is None: --------- loading vocab... 19538 0 Traceback (most recent call last): File "load_embeddings.py", line 24, in d =...

请问对source img有额外的预处理么？或者是训练过程加了额外的crop之类的数据增广？ （因为测试过程发现source img采用不同方式crop,drive的结果差不多。这是怎么做到的呢？）

File "convert_2_nonbnn.py", line 121, in get_netparameter pb.Parse(f.read(), net) AttributeError: 'module' object has no attribute 'Parse' protobuf2.5.0 is there other function like Parse?

conf_center_pred_.count() and bottom[4]->count

4

I want use the loss on https://github.com/zeusees/FaceBoxes/blob/master/examples/faceboxes/train.prototxt I concat conv2_3/incep , conv3_2 , conv4_2 as "mbox_center_conf" center_features: 23 but, the error is: multibox_center_loss_layer.cpp:299] Check failed: conf_center_pred_.count() == bottom[4]->count() (16062464 vs....

preImgOps里把图像resize到256，不是会影响图像原始的清晰度么？

batchnorm

1

why set different param batchnorm, in base layer: nn.BatchNorm2d(out_planes) but in IDAUp and MobileNetUp: nn.BatchNorm2d(out_dim,eps=0.001,momentum=0.1)

is the model same as paper centerface? because I test your model cpu forward time =2s but ,centerface github code cpu forward time=70ms input is 800*800

if tscore < maxscore:

1

请问下边的代码是什么意思呢？ sc不是按大到小排序的，这行代码不会执行呀？

NameError: name 'SCDataset' is not defined

1

CUDA_VISIBLE_DEVICES=0 python train.py --tag mt --model REC

imaging_quality

1

what's the process about 'imaging quality'. (how to eval imaging quality dimension about t2v model) cannot find any code call imaging_quality.py compute_imaging_quality and also cannot find prompt txt about imaging...