sirius-ai MobileFaceNet_TF issues

sirius-ai / MobileFaceNet_TF

Tensorflow implementation for MobileFaceNet

Apache License 2.0

434 stars 170 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Liveness detection

#88 YasseerCP opened 1 year ago
0
使用作者提供的预训练模型参数embeeding（192）和 cls num（85742）必须和作者的相同才能跑起来，否则会报错，有解决方法吗？

#87 howell0203 opened 1 year ago
0
what is step to train tflite custom model please help us to build using own datasets.

#86 ashrafulyarali opened 2 years ago
0
想问一下那个pretrained_model该怎么使用呢我在train_nets.py里面加上了它的路径然后执行报错了不添加pretrained_model时候执行正常但是准确率感人......

#85 DongONNS opened 3 years ago
2
Performance issues in your project (by P3)

#84 DLPerf opened 3 years ago
1
Performance issues in train_nets.py(P2)

#83 DLPerf opened 3 years ago
0
What should be the similarity formula between 2 embeddings (euclidian distance or cosine similarity ) ?

#82 radha1990kumari opened 3 years ago
0
What should be threshold for Euclidean distance between 2 embeddings ?

#81 radha1990kumari opened 3 years ago
2
自己制作的rec文件生成tfrecord出错

#80 masbbx opened 3 years ago
0
bug fix: validation result

#79 dangnh0611 opened 3 years ago
0
Trying to read ckpt but get ValueError:

#78 neesetifa closed 3 years ago
1
embedding没有度量性问题（相同人不同人的embedding距离无差别）

#77 wensi2690 opened 3 years ago
0
convert pb model to tensorflow.js model

#76 shuxiongwuziqi opened 3 years ago
2
convert pb to tflite error

#75 guker opened 3 years ago
0
is refined ms-celeb-v1 dataset is from insightface?

#74 ohjoon2 opened 3 years ago
0
Cannot parse file b'/content/drive/MyDrive/MobileFaceNet/MobileFaceNet_TF-master/arch/pretrained_model/saved_model.pb': Error parsing message.

#73 lnghia opened 3 years ago
7
fix validation and test results not correct

#72 dangnh0611 closed 3 years ago
0
question: input normalization and expected euclidean distance ranges

#71 vladmandic closed 3 years ago
1
你好，请问该如何把test_nets.py运行起来？

#70 qxzhou1010 opened 4 years ago
2
Cannot feed value of shape (112, 112, 3) for Tensor 'input:0', which has shape '(?, 112, 112, 3)'

#69 JionBin closed 4 years ago
1
How to create tfrecords for VGGFace2 dataset using utils/dataprocess.py.

#68 Mamtesh11 opened 4 years ago
2
Can it run on ESP-CAM with tensorflow lite?

#67 manjrekarom opened 4 years ago
5
inputs = tf.placeholder(name='img_inputs', shape=[None, *args.image_size, 3], dtype=tf.float32)

#66 Casi11as opened 4 years ago
6
模型泛化性问题

#65 Dengxianshuai opened 4 years ago
3
Why use this ？ should i use this when inference ？

#64 tomjerrygithub opened 4 years ago
0
Why are ckpt graph output and pb output different?

#63 kimnamu opened 4 years ago
1
在i7-7700上跑提取特征值需要400ms

#62 nsyncxy opened 4 years ago
0
这个thresholds = np.arange(0, 4, 0.01) 是做什么用的？

#61 tomjerrygithub opened 4 years ago
0
请问作者预训练模型用的什么数据集？

#60 jiamingzhang94 closed 3 years ago
3
error when i load the pretrained model

#59 zfs1993 closed 4 years ago
0
请问：用pretrained的模型，为什么验证的精度很高99%，而训练的ms1m的精度只有0.02

#58 happymonkey1982 opened 4 years ago
2
loss不下降

#57 Dengxianshuai opened 4 years ago
3
如何比对两张图片是否是一个人

#56 syaringan357 closed 4 years ago
6
Minor proposals

#55 kant closed 5 years ago
0
损失函数的值从10000轮左右开始就在20左右晃一直不下降了

#54 flyduck closed 4 years ago
19
为什么报错了 Process finished with exit code -1073741819 (0xC0000005)

#53 1374207430 closed 4 years ago
2
it seems the pb file is only graph info

#52 QuarK30 closed 5 years ago
2
Training classification accuracy?

#51 capilano closed 5 years ago
5
how to use the verification.py ? and what the parameters in the functions representing???

#50 zhiweiguo opened 5 years ago
3
作者你好，请问程序中提供了前向计算提取人脸图片深度特征的接口吗

#49 ubuntuyumiao closed 4 years ago
1
作者您好，我的total_loss一直在15~40之间跳，reg_loss倒是一直降低，但降低到0.04的时候就停止了

#48 frankqianghe closed 5 years ago
7
Does not generalise to other face alignments

#47 deepak-neva opened 5 years ago
1
How to convert pretrained model to tflite

#46 fanqie03 closed 4 years ago
21
inference loss

#45 DietDietDiet closed 5 years ago
5
how to test and evaluate it ?

#44 fanqie03 closed 5 years ago
5
1:1 face matching

#43 dear-john closed 5 years ago
7
Adapt for 224x224 input

#42 batrlatom closed 5 years ago
1
> just keep running

#41 CHANGErIBE closed 5 years ago
1
epoch and training accuracy are always 0

#40 jolinlinlin closed 5 years ago
7
accuracy is 0

#39 NicoleNicoleXiang closed 5 years ago
4