Issues
- 30
- 0
Got the ModuleNotFoundError error,please check
#54 opened by qyum - 0
RuntimeError: image.is_contiguous()INTERNAL ASSERT FAILED at ".\\RoIAlign\\roi_align\\src\\crop_and_resize_gpu.cpp":27, please report a bug to PyTorch. image must be contiguous
#53 opened by Kev1n3zz - 4
output roi all zeros
#25 opened by qchenclaire - 0
- 2
- 2
‘’Legacy autograd function with non-static forward method is deprecated.” on Pytorch 1.5
#41 opened by Lucky666123 - 2
- 0
- 1
from ._crop_and_resize import lib as _lib, ffi as _ffi ModuleNotFoundError: No module named 'roi_align._ext.crop_and_resize._crop_and_resize'
#35 opened by KainingYing - 1
Can this function work on pytorch1.5 ?
#40 opened by Lucky666123 - 3
- 0
The name should be crop_and_resize_backward ranther than crop_and_resize_forward?
#47 opened by wsh122333 - 1
ImportError: No module named crop_and_resize_cpu
#45 opened by vusdo - 2
- 3
RuntimeError: cuda runtime error (77) : an illegal memory access was encountered at torch/csrc/cuda/Module.cpp:267
#10 opened by GeoffreyChen777 - 0
Install log
#44 opened by fjssharpsword - 0
- 3
Problem on box_ind
#14 opened by hguosc - 1
Can you provide a method for extracting RoI feature maps of irregular polygon with sixteen vertices?
#39 opened by ustczhouyu - 0
兄弟,你这没按定义吧
#38 opened by sunhao12121 - 5
- 0
crop_and_resize.obj : error LNK2001: unresolved external symbol __imp_THFloatTensor_zero
#37 opened by Yinhance - 0
CropAndResize seems to have wrong outputs
#36 opened by littlespray - 6
- 10
install error
#11 opened by dami23 - 0
fp16 support
#33 opened by turboxin - 1
- 0
fail to install in window10
#31 opened by NaifahNurya - 16
Hi gradcheck failed
#1 opened by mks0601 - 0
import problem
#30 opened by joshua23333 - 3
- 2
PyTorch 1.0.0 support
#29 opened by jjldr - 2
TypeError from crop_and_resize.py : "initializer for ctype 'struct THCudaTensor *' must be a pointer to same type, not cdata 'struct THFloatTensor *"
#26 opened by kathydo - 0
- 0
Crop height and Crop width
#23 opened by medhini - 0
adding nms to this package
#22 opened by IssamLaradji - 3
pytorch-1.0
#20 opened by fabienbaradel - 1
pytorch1.0 ffi was deprecated
#21 opened by xiaoshihoua - 0
- 1
why can not use float64
#18 opened by wu-zhonghua - 2
crop a patch from image
#15 opened by liuqk3 - 1
It supports batchsize>1?
#16 opened by gjd2017 - 0
what's the difference of this method to first resizing the feature map to original image size
#7 opened by ypxie - 0
How about the GPU memory cost?
#6 opened by sshaoshuai - 2
error with test.py
#5 opened by xpain1 - 2
GPU support
#3 opened by simo23