xuannianz/keras-CenterNet
CenterNet (Objects as Points) implementation in Keras and Tensorflow
PythonApache-2.0
Issues
- 0
Centernet bbox height width label
#65 opened by soans1994 - 1
- 3
I run the code with tensorflow 2.1.0
#27 opened by m1nt07 - 4
Try to train by COCO dataset
#45 opened by vladimircape - 1
他的代码有问题
#46 opened by Salary-only-17k - 0
Same issue as 33 not able to train on custom data
#44 opened by rnb007 - 4
Max_Objects > 100 will raise error
#30 opened by chuong-phung - 2
- 1
- 1
mAP is 0
#43 opened by linchunmian - 1
- 1
About subsampling and maxpooling
#40 opened by CedarYang - 2
ValueError: Shape must be rank 4 but is rank 3 for 'lambda_1/concat' (op: 'ConcatV2') with input shapes:
#24 opened by chenxiao521 - 0
- 1
- 1
- 1
about:preprocess_image
#36 opened by CedarYang - 3
mAP is always zero on COCO dataset
#19 opened by IaroslavS - 3
Do you have the issue like first img inference slow on your tf or keras tf backend?
#31 opened by AutumnZ-94 - 1
mAP of resnet50
#32 opened by kp-1 - 1
- 1
Can't sve my model?why
#34 opened by cuiziteng - 1
About the cpu version of CenterNet
#29 opened by chenzhengnan - 0
Why does the eval code flip image?
#28 opened by MioChiu - 2
- 12
voc resnet50 mAP=0.7290
#15 opened by zgsxwsdxg - 1
Reg_mask_input
#21 opened by GreatGameDota - 1
Train own dataset
#23 opened by DLH06 - 1
Very slow training
#20 opened by ARGSK - 2
how about tensorflow version?
#17 opened by zcunyi - 1
- 6
No module named 'utils.compute_overlap'
#13 opened by apardogi - 1
keras-CenterNet-master/losses.py", line 28, in reg_l1_loss: gather() got an unexpected keyword argument 'batch_dims'
#14 opened by lyy-fighting - 2
Convert pytorch pre-trained model to keras
#11 opened by WangyiNTU - 2
- 2
freeze_to_pb
#7 opened by xianglizhi123 - 4
Resnet18 backbone
#10 opened by ARGSK - 2
No boxes problem
#4 opened by AirFishWang - 3
about pretrained model
#3 opened by AirFishWang - 1
License
#12 opened by DrFirestream - 2
train/loss
#6 opened by JayjieL - 3
No module named 'keras_resnet'
#2 opened by ARGSK - 1
load model?
#8 opened by AirFishWang - 3
the Loss not converging
#5 opened by gugugilin - 1
how can i get the Pretrained model
#1 opened by ARGSK