instance-segmentation-pytorch
instance-segmentation-pytorch copied to clipboard
Wizaron
Semantic Instance Segmentation with a Discriminative Loss Function in PyTorch
Hi @Wizaron I have the dependency problems when creating conda env: `conda env create -f instance-segmentation-pytorch/code/conda_environment.yml` 1/ Conda supports cudatoolkit >= 9.0 so I can't install `cudatoolkit==8.0` ``` Collecting package...
Traceback (most recent call last): File "/home/hongrui/anaconda3/lib/python3.8/site-packages/torch/multiprocessing/spawn.py", line 59, in _wrap fn(i, *args) File "/home/hongrui/project/metro_pro/instance-segmentation-pytorch/code/train_engine.py", line 60, in main_worker worker.validation(epoch) File "/home/hongrui/project/metro_pro/instance-segmentation-pytorch/code/train_worker.py", line 219, in validation mb_out_metrics, loss, outputs =...
hi, how can I use tsne in a semantic segmentation network to visualize high-dimensional features? actually, each pixel has a vector embedding, am I right? especially I don't know how...
If training with **usegpu=False** , will throw assert error of size_average.
the nums in the array n_clusters involves background?
In the implementation of 'recurrent_hourglass.py'. There is only one```ConvGRUCell``` ``` self.convgru_cell = ConvGRUCell ``` But it seems that there are supposed to be ```n_levels ConvGRUCell```, each of which tackles one...
There are too many data directories about CVPP, a universe training dataset preparation would better to understand whole process training
Does the code provide the Average Precision evaluation metric for cityscapes ? Thanks!
