Issues
- 2
- 4
unable to install ./init.sh error
#149 opened by bat3a - 1
[HELP] problem with cpu_nms
#154 opened by m0rp43us - 2
requirements.txt
#131 opened by crystal786 - 3
In-depth guide to getting FCIS working on Windows 10
#153 opened by ferakon - 0
Result file for COCO test-dev instance segmentation
#166 opened by RbtJhs - 3
pip install -r requirements.txt
#148 opened by Kobeee0614 - 4
- 2
Trouble building on Windows 10 (TypeError: expected str, bytes or os.PathLike object, not NoneType)
#152 opened by ferakon - 0
Is it possible to run FCIS on AMD?
#151 opened by vincent-26 - 0
Problems when launching demo.py
#150 opened by vincent-26 - 1
- 3
有哪位遇到过训练成功,测试map=0的情况吗
#146 opened by ccxietao - 5
When I train my own coco-like datasets,I successfully train and test the model,but the result of all AP is 0.
#143 opened by CPFelix - 3
- 1
mxnet.base.MXNetError: [16:31:18] src/operator/nn/./cudnn/cudnn_convolution-inl.h:449: Check failed: e == CUDNN_STATUS_SUCCESS (3 vs. 0) cuDNN: CUDNN_STATUS_BAD_PARAM
#147 opened by CPFelix - 0
my own data set ,test map is 0
#145 opened by ccxietao - 1
When I train on Pascal VOC the result of mAP is 0
#112 opened by PerthyChen - 1
- 0
Network structure
#142 opened by yyangmei - 0
The result finetune with cityscapes is very poor
#141 opened by vtHuo - 1
mxnet.base.MXNetError: [15:20:32] src/engine/threaded_engine.cc:320: Check failed: device_count_ > 0 (-1 vs. 0) GPU usage requires at least 1 GPU
#140 opened by zhanghui-hunan - 2
How can I train FCIS with cityscapes dataset?
#125 opened by ChangliangSun - 0
Some problems encountered when using deconvolution
#138 opened by sifengyang - 1
TypeError: _update_params_on_kvstore() takes exactly 4 arguments (3 given)
#135 opened by Danee-wawawa - 1
cannot import name bbox_overlaps_cython
#132 opened by wenyaole - 1
- 0
- 1
OpenCV Error: Assertion failed (dsize.area() > 0 || (inv_scale_x > 0 && inv_scale_y > 0))
#113 opened by PerthyChen - 0
some questions about make Mxnet on my computer .
#133 opened by YingXiuHe - 1
No bias for each convolutional layer? Why?
#128 opened by zhaodongsun - 3
Slightly Lower Accurancy Using One GPU
#130 opened by zhaodongsun - 0
- 0
About ChannelOperator.
#121 opened by franciszzj - 0
Speed of training
#127 opened by michaelhuang74 - 0
Can't download resnet-v1-101.params, anyone can help me to download this file?
#126 opened by ChangliangSun - 0
How can I train from specific Epoch?
#124 opened by hlthu - 5
- 3
Missing files after running .\init.bat
#120 opened by SunXz - 0
- 1
GPU memory usage.
#119 opened by franciszzj - 0
代码很屌,但是论文写的什么玩意儿?
#108 opened by yelusaleng - 1
- 0
demo.py no masks on custom dataset, works on coco.
#115 opened by Matimath - 0
AssertionError: shape inconsistent for fcis_cls_seg_weight inferred when using own dataset
#114 opened by NikTRSK - 0
Can I transport the mask results into ploygon?
#109 opened by zouxinh - 0
cblas.h: No such file or directory ??
#110 opened by MachineLP - 0
- 0
that's a brilliant job!!!
#104 opened by jacktkk - 0
Performance of pre-trained model (Demo)
#103 opened by ld-ing