luckycallor InsightFace-tensorflow issues

luckycallor / InsightFace-tensorflow

Tensoflow implementation of InsightFace (ArcFace: Additive Angular Margin Loss for Deep Face Recognition).

248 stars 112 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

embedding comparision

#48 marziehoghbaie opened 3 years ago
0
请问一下不同数据集训练一些超参数怎么设置的？

#47 Nakupenda-7 opened 4 years ago
0
训练时Acc会一直为0吗？

#46 zch-ssc opened 4 years ago
4
请问backbone里的resnet_v1和v2以及modifiedv2有对比吗，哪个效果最好？

#45 FelixZhang7 opened 4 years ago
0
关于训练时loss的问题

#44 vvblack opened 4 years ago
1
下载lfw官网数据集制成tfrecord后运行finetune_softmax.py时acc起始很低

#43 chenkeyi2013 opened 4 years ago
3
questions about training dataset

#42 vvblack opened 4 years ago
1
questions about threshold

#41 vvblack opened 4 years ago
1
good job

#40 damaocun opened 4 years ago
1
output node of the second pre-model(from baidu Drive, ａs if it were the first pre-trained model

#39 tjulitianyi1997 opened 4 years ago
1
经测试效果不好，本项目还需要进一步完善！

#38 CachCheng opened 4 years ago
0
When I want to use my own data to train the model, I first transfer the dataset to .tfrecord format. How can I set the validation data format in config file? Just use the .tfrecord or only use .bin?

#37 GodMon opened 4 years ago
4
多GPU训练，获取embedding的时候报错KeyError: "The name 'embeddings:0' refers to a Tensor which does not exist.

#36 HandsLing opened 4 years ago
0
Model for any other case is severely overfitted

#35 AkhilAshref opened 4 years ago
0
Ask about IDE

#34 ThanhPham1987 opened 4 years ago
1
FileNotFoundError:

#33 snow-wind-001 opened 4 years ago
1
Couple of errors in get_embd.py

#32 AkhilAshref opened 4 years ago
0
missing requirements

#31 SaltyHash123 opened 4 years ago
0
Extracting incorrect embedding in get_embd.py

#30 qwn19970728 opened 5 years ago
0
两张照片不同的人，距离也很小

#29 HandsLing opened 5 years ago
2
mode of mxrec convert to tfrecord

#28 Deepcong2019 opened 5 years ago
3
re-implementation with tensorflow cannot achieve the same accuracy on mxnet.

#27 Jar7 opened 5 years ago
1
1006k的预训练模型和334k的好像是同样一个压缩包

#26 sunruina2 opened 5 years ago
2
should not only restore trainable variables in `get_embd.py`?

#25 HongtaoYang opened 5 years ago
1
Normalization of embeddings in evaluate.py

#24 AmitRozner opened 5 years ago
0
Questions about training

#23 christk1 closed 5 years ago
2
使用训练好的模型在faces_ms1m_112x112.tfrecord数据上验证时准确率很低？

#22 xuweihuawei opened 5 years ago
4
ValueError: dictionary update sequence element #0 has length 5; 2 is required

#21 2151330 opened 5 years ago
1
Question: How do I generate a validation set .bin file with my own data?

#20 kingguan4 opened 5 years ago
1
Question about: "face images should be well cropped here"

#19 khanhnt opened 5 years ago
2
Why do you have configurations of different batch sizes?

#18 BruceDai003 opened 5 years ago
1
Why you dont add summary to summary writer?

#17 milk-bottle-liyu opened 5 years ago
0
How do I validate on my own validation set?

#16 Wenlin-Chen opened 5 years ago
0
2 different faces but small distance?

#15 huytq97 opened 5 years ago
18
finetune loss always big

#14 x3393467 opened 5 years ago
2
Is there any problem about get_embd.py？

#13 YJingyu opened 5 years ago
3
训练模型的准确率

#12 destiny311 opened 5 years ago
3
pre_trained models

#11 destiny311 opened 5 years ago
1
How exactly is the cropping done?

#10 lagidigu opened 5 years ago
5
distance_metric == 1,evaluate.py, when to choose 'distance_metric == 1'?

#9 Wxsang opened 5 years ago
2
how to compare to embeddins of faces for knowing that these faces are similar

#8 Muhammad333 opened 5 years ago
2
support specify visible gpus

#7 jasstionzyf opened 5 years ago
3
questins about training time

#6 friedhelm739 closed 5 years ago
4
checkpoint file so large(643M)

#5 clhne opened 5 years ago
5
About image normalization

#4 yinguobing closed 5 years ago
2
Low accuracy

#3 nnt296 closed 5 years ago
1
valid dataset

#2 csqzs opened 5 years ago
2
Pre-trained models

#1 TimoBolkart closed 5 years ago
3