improving_segmentation_with_selfsupervised_depth
improving_segmentation_with_selfsupervised_depth copied to clipboard
lhoyer
[CVPR21] Implementation of our work "Three Ways to Improve Semantic Segmentation with Self-Supervised Depth Estimation"
Hello @lhoyer I got some errors when I try to use my own images. (threeWays) D:\Work\improving_segmentation_with_selfsupervised_depth-master>python inference.py --machine ws --model model --data t Start inference2022-02-11_11-01-23-734969 LOG_DIR RUNDIR: results/\inference2022-02-11_11-01-23-734969/ C:\Anaconda\envs\threeWays\lib\site-packages\torchvision\transforms\transforms.py:287: UserWarning:...
I see your datasets need [leftImg8bit_sequence_trainvaltest.zip (324GB)](https://www.cityscapes-dataset.com/file-handling/?packageID=14) [[md5]](https://www.cityscapes-dataset.com/md5-sum/?packageID=14),it's too big. I think my machine may spend long time to train a model. So I want to know your machine use...
Hi, i want to know what is the input if i use inference.py to test my own image, is only a RGB image or RGB image and its depth image?...
Hi, when i run python inference.py --machine ws --model /path/to/checkpoint/dir/ --data /path/to/data/dir/, the erro is: zipfile.BadZipFile: File is not a zip file. I found that is the mono_cityscapes_1024x512_r101dil_aspp_dec6_lr5_fd2_crop512x512bs4.zip cant unzip,...
hello， i'm very interested in the paper. I have some questions about the execution of the entire process. 1."python train.py --machine ws --config configs/cityscapes_monodepth_highres_dec5_crop.yml python train.py --machine ws --config configs/cityscapes_monodepth_highres_dec6_crop.yml"...
