Computer vision

Hi, Mr.Wang Happy Chinese New Year and wish all the best to you! B.R Anguo

hi @huangyangyu 非常感谢。英文打字比较累，还是说中文吧，希望不要介意：） 另外，还有一个问题，我在windows/CPU编译了一下代码，运行也没问题，然后提取特征的时候，发现fc5的512个float值有些非常大，比如317.xxx之类的，然后我通过cos 计算两个array的相似度的时候，任意两张人脸的相似度都是99%，很奇怪，然后我看了一下代码： def cos_sim(v1, v2): dist = np.linalg.norm(np.array(v1)-np.array(v2)) cos = 1 - dist * dist / 2 return cos 这个貌似不是cos的计算公式吧

另外，我看到这个代码，貌似是在给每个float值做标准化？ def get_feature(featurer, image_dir, feature_dim=512): final_feature = np.zeros(feature_dim) for name in os.listdir(image_dir): image_file = image_dir + name feature = featurer.test(image_file=image_file) if feature is not None: final_feature += feature / np.linalg.norm(feature,...

Thank you for your response, but I really puzzled on the output of fc5 layer, because when I calculate the cos sim between two float vectors, I always get the...

I also tried to normalize the output of fc5, this is what I have done: 1.compiled your source code on windows, successfully 2.modified and compiled the classify demo in C++...

Hi@wuyongchn , where is the image list, it seems there is no such files or dirs in this repository

yes, I changed it into other resolution, it was fine(which means no errors, the result size is correct), but the scale ratio seems not correct, even use your default 480x640,...

@Howave the model you downloaded is 640x640 input(from the caffe prototxt, but I think the input should same with mxnet one), however, in your source code, it is 480(width)x640(height), have...

with the source code: model_irse.py, I could convert Backbone_IR_152_Epoch_112_Batch_2547328_Time_2019-07-13-02-59_checkpoint.pth (Converted successfully into .pt and load it in C++) but still with problems on: Head_ArcFace_Epoch_112_Batch_2547328_Time_2019-07-13-02-59_checkpoint.pth Could you please give some advice?...

Message=Error(s) in loading state_dict for Backbone: Missing key(s) in state_dict: "input_layer.0.weight", "input_layer.1.weight", "input_layer.1.bias", "input_layer.1.running_mean", "input_layer.1.running_var", "input_layer.2.weight", "output_layer.0.weight", "output_layer.0.bias", "output_layer.0.running_mean", "output_layer.0.running_var", "output_layer.3.weight", "output_layer.3.bias", "output_layer.4.weight", "output_layer.4.bias", "output_layer.4.running_mean", "output_layer.4.running_var", "body.0.res_layer.0.weight", "body.0.res_layer.0.bias", "body.0.res_layer.0.running_mean", "body.0.res_layer.0.running_var",...