GLSAN

GLSAN is a network for drone-view small object detection.

Installation

Our source codes are mainly based on Detectron2, see Detectron2.installation.

Get Started

About the initialization of Detectron2, please refer to Detectron2.Getting_started.

dataset transformation

To train the VisDrone and UAVDT dataset, you need transform them to coco format. We provide './tools/txt2xml_*.py' and './tools/xml2json_*.py' to generate json files in coco format.

dataset augmentation

The network in our paper is trained with the augmented datasets. We provide './tools/crop_dataset.py' and './tools/sr_dataset.py' to conduct SARSA and LSRN to the original datasets.

pretrained models

The pretrained models of our network can be downloaded at Detectron2.model_zoo. You can directly download R-50.pkl or R-101.pkl to '.torch/fvcore_cache/detectron2/ImageNetPretrained/MSRA/' of your 'home' directory. Or they will be downloaded automatically when training.

training

We provide "train_net.py" for network training. To train a model with "train_net.py", first setup the corresponding datasets following Detectron2.datasets, you need to put the transformed or augmented datasets into './datasets' directory. The settings of VisDrone and UAVDT can be found in './glsan/data/datasets'.

To train with 8 GPUs, run:

python train_net.py --config-file ./configs/faster_rcnn_res50_visdrone.yaml --num-gpus 8

To train with 1 GPU, run:

python train_net.py --config-file ./configs/faster_rcnn_res50_visdrone.yaml --num-gpus 1 SOLVER.IMS_PER_BATCH 2

evaluation

To evaluate a model's performance, there are threee modes corresponding to three different cropping strategies: NoCrop, UniformlyCrop, SelfAdaptiveCrop. You can run following codes to switch the cropping strategy:

python train_net.py --config-file ./configs/faster_rcnn_res50_visdrone.yaml --eval-only --num-gpus 8
python train_net.py --config-file ./configs/faster_rcnn_res50_visdrone.yaml --eval-only --num-gpus 8 GLSAN.CROP UniformlyCrop
python train_net.py --config-file ./configs/faster_rcnn_res50_visdrone.yaml --eval-only --num-gpus 8 GLSAN.CROP SelfAdaptiveCrop

To add super-resolution operation to the network, run:

python train_net.py --config-file ./configs/faster_rcnn_res50_visdrone.yaml --eval-only --num-gpus 8 GLSAN.CROP SelfAdaptiveCrop GLSAN.SR True

To acquire more parameters of our method, see './glsan/config/defaults.py' and './glsan/modeling/meta_arch/glsan.py'

dengsutao / glsan