Super-Resolution Assisted Dual-Branch YOLO for Enhanced Small Target Detection in Remote Sensing Images
Welcome to the repository for Super-Resolution Assisted Dual-Branch YOLO (ASR-YOLO), an advanced algorithm aimed at balancing image resolution and computational resource consumption to enhance its real-time capabilities in small target detection in remote sensing images.
The code in this repository provides data support for the paper "Super-Resolution Assisted Dual-Branch YOLO for Enhanced Small Target Detection in Remote Sensing Images".
base ----------------------------------------
matplotlib>=3.2.2 numpy>=1.18.5 opencv-python>=4.1.2 Pillow PyYAML>=5.3.1 scipy>=1.4.1 torch>=1.7.0 torchvision>=0.8.1 tqdm>=4.41.0
logging -------------------------------------
tensorboard>=2.4.1 wandb
plotting ------------------------------------
seaborn>=0.11.0 pandas
export --------------------------------------
coremltools>=4.1 onnx>=1.8.1 scikit-learn==0.19.2 # for coreml quantization
extras --------------------------------------
thop==0.0.31.post2005241907 # FLOPS computation pycocotools>=2.0 # COCO mAP
results--------------------------------------
xlsxwriter>=3.0.1
Download datasets from the baiduyun (code: hvi4) links and place them in this directory.
transform_vedai.py:Dataset Processing
test.py:Inference and test of the ASR_YOLO Model
train.py:Training the ASR_YOLO Model
models:Store the source code of models for comparison.