Skip to content

nuoxu/VLAS

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

28 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Visual-Language Active Search for Wide-Area Remote Sensing Imagery

Setup

  • You should install prerequisites using: pip install -r requirements.txt
  • Our settings follow previous works, please refer to VAS.

Dataset

Please downoad DOTA and xView and place them in dataset/. Prepare annotation files for the two datasets:

python3 tools/prepare_data_for_dota.py
python3 tools/prepare_data_for_xview.py

Training and Evaluation

Training

python3 tools/train-VLAS.py \
      --dataset DOTA (or xView) \
      --cv_dir path_to_you_model \
      --num_actions 36/49/64/81/100 \
      --num_cluster 12/24/36 \
      (--multiclass)
python3 tools/train-PAGE.py \
      --dataset DOTA \
      --cv_dir path_to_you_model \
      --num_actions 100

Evaluation

python3 tools/test-VLAS.py \
      --dataset DOTA \
      --cv_dir path_to_you_model \
      --num_actions 100
python3 tools/test-PAGE.py \
      --dataset DOTA \
      --cv_dir path_to_you_model \
      --num_actions 100

Visualization

python3 tools/vis-PAGE.py \
      --dataset DOTA \
      --cv_dir path_to_you_model \
      --num_actions 100

About

Visual-Language Active Search for Wide-Area Remote Sensing Imagery (Pattern Recognition 2026)

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages