Are you sure you want to delete this task? Once this task is deleted, it cannot be recovered.
unknown de6c45d8ba | 1 year ago | |
---|---|---|
.. | ||
README.md | 1 year ago | |
hccl_4p_0123_127.0.0.1.json | 1 year ago | |
hccl_tools.py | 1 year ago | |
run_distribute_train.sh | 1 year ago | |
run_infer_310.sh | 1 year ago | |
run_single_eval.sh | 1 year ago | |
run_single_train.sh | 1 year ago | |
run_single_train_shape.sh | 1 year ago |
git clone https://github.com/Walstruzz/edge_eval_python.git
cd edge_eval_python/
cp -R impl ../src
cp eval_edge.py ../src
cp nms_process.py ../src
in edges_eval_plot.py:
delete line 71-73 and 78
in edges_eval_dir.py:
change line 10: from src.impl.bwmorph_thin import bwmorph_thin
change line 11: from src.impl.correspond_pixels import correspond_pixels
after line 50(for g in gt:) add:
h, w = g.shape
if h > w:
g = np.flip(g,axis=1)
g = np.transpose(g)
after line 78(for _g in _gt:) add:
h, w = _g.shape
if h > w:
_g = np.flip(_g,axis=1)
_g = np.transpose(_g)
delete line 105-107(from if to return) and add:
np.savetxt(out, info, fmt="%10g")
in nms_process.py:
add in line 5: from src.model_utils.config import config
change line 6: from src.impl.toolbox import conv_tri, grad2
change line 16: solver = cdll.LoadLibrary(os.path.join(config.cxx_path, "lib/solve_csa.so"))
change line 53: assert file_format in {".mat", ".npy", ".bin"}
after line 72(image = np.load(abs_path)) add:
elif file_format == ".bin":
image = np.fromfile(abs_path, dtype=np.float32).reshape(321, 481)
in eval_edge.py:
change line 4: from src.impl.edges_eval_dir import edges_eval_dir
change line 5: from src.impl.edges_eval_plot import edges_eval_plot
in correspond_pixels.py:
add in line 5: from src.model_utils.config import config
add in line 10: import os
change line 11: solver = cdll.LoadLibrary(os.path.join(config.cxx_path, "lib/solve_csa.so"))
cp -R cxx ../src
Dear OpenI User
Thank you for your continuous support to the Openl Qizhi Community AI Collaboration Platform. In order to protect your usage rights and ensure network security, we updated the Openl Qizhi Community AI Collaboration Platform Usage Agreement in January 2024. The updated agreement specifies that users are prohibited from using intranet penetration tools. After you click "Agree and continue", you can continue to use our services. Thank you for your cooperation and understanding.
For more agreement content, please refer to the《Openl Qizhi Community AI Collaboration Platform Usage Agreement》