minghanqin
Problems encountered when processing my own scene
Dear Author，Thank you for open-source such a great work Following README.md,I have successfully completed the configuration of the environment. But problems occoured when I tried to process my own scene:
Step 2: Train the Autoencoder and get the lower-dims Feature.
#train the autoencoder
cd autoencoder
python train.py --dataset_name $dataset_path --encoder_dims 256 128 64 32 3 --decoder_dims 16 32 64 128 256 256 512 --lr 0.0007 --output ae_ckpt
#get the 3-dims language feature of the scene
python test.py --dataset_name $dataset_path --output
Firstly,there seems to be issues with the command you give:
python train.py --dataset_name $dataset_path --encoder_dims 256 128 64 32 3 --decoder_dims 16 32 64 128 256 256 512 --lr 0.0007 --output ae_ckpt
When I enter this command in the terminal, the following error message appears:
So I modified this command to:
python train.py --dataset_path ./dataset/413/input/ --encoder_dims 256 128 64 32 3 --decoder_dims 16 32 64 128 256 256 512 --lr 0.0007 --dataset_name 413（But I'm not sure if this modification is correct.）
After running this command,New error messages have appeared：
It seems the issue is in the dataset.py file.
！！！Has anyone encountered this issue before? How should it be resolved?
Any help would be appreciated！！！
I found a solution, haven't tested further but so far it's working
on autoencoder/dataset.py
change data_names variable from
data_names = glob.glob(os.path.join(data_dir, '*f.npy'))
to
data_names = glob.glob(os.path.join('./..',data_dir, '*f.npy'))
the error is because the directory path to language features is incorrect as we do cd autoencoder
hope this helps
