tianzhi0549 FCOS issues

tianzhi0549 / FCOS

FCOS: Fully Convolutional One-Stage Object Detection (ICCV'19)

https://arxiv.org/abs/1904.01355

Other

3.28k stars 630 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

my own coco datasets have no mask annotations,how to turn off the mask??

#145 dreamhighchina closed 5 years ago
0
import error

#144 YYZ-rose opened 5 years ago
1
compile error

#143 dreamhighchina closed 5 years ago
1
import error

#142 dreamhighchina closed 5 years ago
1
Inference speed on MobileNetV2

#141 stigma0617 closed 5 years ago
5
Inference speed on MobileNetV2

#140 stigma0617 closed 5 years ago
0
pos_inds number bug in loss function in fcos

#139 Baby47 closed 5 years ago
1
Some questions about the results in Table 4 in the paper

#138 chensnathan closed 5 years ago
2
How to test on coco

#137 hheavenknowss opened 5 years ago
4
IOU loss or Smooth L1 loss?

#136 BlueBlueFF opened 5 years ago
2
can i train the retinanet with your code？

#135 qianzhang2018 opened 5 years ago
1
AP of FCOS is lower than RetinaNet?

#134 boundles opened 5 years ago
1
A problem during inference on my own datasets

#133 stellating opened 5 years ago
12
The performance is very bad and disappointed when using fcos

#132 alexwq100 opened 5 years ago
12
How to train images without targets

#131 adizhol closed 5 years ago
2
train stopped

#130 mama110 closed 5 years ago
2
Testing time in c++ is 10 times than pytorch ！

#129 xscjun closed 5 years ago
1
RuntimeError: CUDA error: device-side assert triggered

#128 mama110 closed 5 years ago
4
I train the net on coco2017, but get a problem on transformer Resize

#127 cunjif closed 5 years ago
2
prepare targets for multiple object

#126 mama110 closed 5 years ago
2
pos_inds

#125 mama110 closed 5 years ago
2
I only have one GPU(GTX1060)

#124 hello-piger opened 5 years ago
6
How can I converting the model to torch script, I failed to do that with 'torch.jit.trace'

#123 xscjun closed 5 years ago
5
How to get l, r, t, b parameters?

#122 SebastienLinker closed 5 years ago
2
Did anybody try FP 16 training?

#121 dedoogong opened 5 years ago
8
How can I change a lighter backbone, such as resnet 18?

#120 Captain-bing opened 5 years ago
1
Trying to train FCOS with different batch size

#119 liyz15 closed 5 years ago
2
subprocess.CalledProcessError: Command '

#118 xuetingdai opened 5 years ago
8
RuntimeError: cuDNN error: CUDNN_STATUS_EXECUTION_FAILED

#117 Captain-bing closed 5 years ago
4
CPU training broken at sigmoid_focal_loss_cpu

#116 kulkarniniraj opened 5 years ago
1
FCOS is too sensitive to batch size?

#115 karlind closed 5 years ago
3
How to transfer your checkpoint to mmdetection framework?

#114 BlueBlueFF opened 5 years ago
3
How does the best possible recalls calculated?

#113 Epiphqny closed 5 years ago
4
grouped_batch_sampler issue : IndexError: index 0 is out of bounds for dimension 0 with size 0

#112 heng2j closed 5 years ago
7
Only one box per location?

#111 lsrock1 closed 5 years ago
1
the question about the positive target number

#110 lihao930104 opened 5 years ago
15
ImportError: cannot import name 'is_pytorch_1_1_0_or_later'

#109 SkeletonOne closed 5 years ago
7
comparison of inference time to Faster R-CNN

#108 stigma0617 closed 5 years ago
13
FCOS_syncbn_bs32_c128_MNV2_FPN_1x model size

#107 Tim5Tang closed 5 years ago
2
Tiny bug in fcos/loss.py

#106 zzzzzz0407 opened 5 years ago
1
Issues running the jupyter Nootbook demo

#105 heng2j closed 5 years ago
1
fix aera typo

#104 hkzhang95 closed 5 years ago
0
is there any tensorflow or caffe project for FCOS

#103 YoungerGao opened 5 years ago
7
FCOS_R_50_FPN_1x mAP 36.59

#102 WitLes closed 5 years ago
12
Hi ， excellent work！ In your paper ,you said the FCOS can replace the RPN in two-stage framework.So how shoulf I change the codes? Could you share your opinions?

#101 runzeer opened 5 years ago
16
Code immediately stop when training starts

#100 lxy598951 closed 5 years ago
2
Did you freeze parts of the backbone when training?

#99 ZhengMengbin closed 5 years ago
2
training error

#98 yokings closed 5 years ago
3
training runtime error

#97 yokings opened 5 years ago
2
about the loss

#96 xuetingdai closed 5 years ago
3

Previous Next