search

JHKim-snu / GVCCI

[IROS 2023] GVCCI: Lifelong Learning of Visual Grounding for Language-Guided Robotic Manipulation
6 stars 0 forks source link

excuse me #2

Open Sunsundayz1 opened 9 months ago

Sunsundayz1 commented 9 months ago

when i run: OMP_NUM_THREADS=4 CUDA_VISIBLE_DEVICES=0,1,2,3 python extract.py --load_dir ./output_caffe152/ --image_dir ../data/train/ENV1_train/ --out_path ../instruction_generation/data/detection_results/ENV1/r152_attr_detection_results --image_list_file ./ENV1_train_train_imagelist_split0.txt --vg_dataset ENV1_train --cuda --split_ind 0 output shows ModuleNotFoundError: No module named 'FasterRCNN'

so I install the bottom-up-attention.pytorch from github, but how can i import it in the project. Look forward your reply

JHKim-snu commented 9 months ago

To address the issue, could you please organize the files and folders under bottom-up-attention.pytorch by placing them under the visual_feature_extraction? This will result in a folder structure, as illustrated below:

|─ visual_feature_extraction
|── FasterRCNN
|── detectron2 |── evaluation |── extract.py ...

Thank you for your cooperation. If you encounter any further challenges or have additional questions, feel free to let us know.

Sunsundayz1 commented 9 months ago

Ok，i put them under my dir But the detectron2 library is a bit wrong, whether it's because the version is wrong 1

from detectron2.structures import BoxMode
  File "/home/zzj/code/gvcci/GVCCI-main/visual_feature_extraction/detectron2/structures/__init__.py", line 6, in <module>
    from .keypoints import Keypoints, heatmaps_to_keypoints
  File "/home/zzj/code/gvcci/GVCCI-main/visual_feature_extraction/detectron2/structures/keypoints.py", line 6, in <module>
    from detectron2.layers import interpolate
  File "/home/zzj/code/gvcci/GVCCI-main/visual_feature_extraction/detectron2/layers/__init__.py", line 3, in <module>
    from .deform_conv import DeformConv, ModulatedDeformConv
  File "/home/zzj/code/gvcci/GVCCI-main/visual_feature_extraction/detectron2/layers/deform_conv.py", line 10, in <module>
    from detectron2 import _C
ImportError: cannot import name '_C' from 'detectron2' (/home/zzj/code/gvcci/GVCCI-main/visual_feature_extraction/detectron2/__init__.py)
JHKim-snu commented 9 months ago

Following the original repo, there was nothing wrong for me.

image

Would you try again with the shared directory?