Shilong Zhang SEPC Issues

jshilong/SEPC Issues

Welcome update to OpenMMLab 2.0
Updated a year ago
I can not find from .version import __version__, short_version, it's a problem
Updated 2 years ago
请问是否有未封装在mmdetection中的版本？
Updated 3 years ago
PConv with Faster RCNN
Closed 3 years ago2
the problem of fcos+sepc
Updated 4 years ago
How do you save the output from each layer of the feature pyramid
Closed 4 years ago1
hello may i ask how you can calculate the correlation matrix of feature maps in fpn
Closed 4 years ago5
Is there a specific reason why we need to set img/gpu >=4 when ibn=True
Closed 4 years ago1
Can't Find SelfRetinaHead
Closed 4 years ago1
NO SelfRetinaHead
Closed 4 years ago1
The question about PConv
Closed 4 years ago3
ATSS head
Closed 4 years ago2
What is the RetinaNet baseline you use in your experiment?
Closed 4 years ago2
Pyramid convolution
Closed 4 years ago2
The evaluation does not work
Closed 4 years ago1
why lconv and cconv in SPEC do not have padding value?
Closed 4 years ago2
How to set each epoch to be evaluated?
Closed 4 years ago1
How about stacked four PConv shared with same parameters?
Closed 4 years ago2
when I train my own model using your configure file and test, error happens.
Closed 4 years ago1
faster rcnn pconv
Closed 4 years ago4
Can sepc_freeanchor be trained with multiple GPUs?
Closed 4 years ago1
How about using Pconv in FPN construction?
Closed 4 years ago2
Flops for deformable conv in PConv
Closed 4 years ago2
Pconv Multi-scale training AP is zero.
Closed 4 years ago3
Have you compared SEPC+FreeAnchor/SEPC+FSAF on X101?
Closed 4 years ago1
Where could I read your paper?
Closed 4 years ago1

jshilong / SEPC

jshilong/SEPC Issues

Welcome update to OpenMMLab 2.0

I can not find from .version import version, short_version, it's a problem

请问是否有未封装在mmdetection中的版本？

PConv with Faster RCNN

the problem of fcos+sepc

How do you save the output from each layer of the feature pyramid

hello may i ask how you can calculate the correlation matrix of feature maps in fpn

Is there a specific reason why we need to set img/gpu >=4 when ibn=True

Can't Find SelfRetinaHead

NO SelfRetinaHead

The question about PConv

ATSS head

What is the RetinaNet baseline you use in your experiment?

Pyramid convolution

The evaluation does not work

why lconv and cconv in SPEC do not have padding value?

How to set each epoch to be evaluated?

How about stacked four PConv shared with same parameters?

when I train my own model using your configure file and test, error happens.

faster rcnn pconv

Can sepc_freeanchor be trained with multiple GPUs?

How about using Pconv in FPN construction?

Flops for deformable conv in PConv

Pconv Multi-scale training AP is zero.

Have you compared SEPC+FreeAnchor/SEPC+FSAF on X101?

Where could I read your paper?