SAPD (Soft Anchor Point Object Detection)

March 13, 2020 ยท View on GitHub

This is an implementation of SAPD for object detection on Keras and Tensorflow. The project is based on fizyr/keras-retinanet, qubvel/efficientnet, xuannianz/EfficientDet and xuannianz/FSAF. The pretrained EfficientNet weights files are downloaded from Callidior/keras-applications/releases

Thanks for their hard work. This project is released under the Apache License. Please take their licenses into consideration too when use this project.

Train

build dataset

  1. Pascal VOC
    • Download VOC2007 and VOC2012, copy all image files from VOC2007 to VOC2012.
    • Append VOC2007 train.txt to VOC2012 trainval.txt.
    • Overwrite VOC2012 val.txt by VOC2007 val.txt.
  2. MSCOCO 2017
    • Download images and annotations of coco 2017
    • Copy all images into datasets/coco/images, all annotations into datasets/coco/annotations
  3. Other types please refer to fizyr/keras-retinanet)

train

  • STEP1: python3 train.py --snapshot imagenet --phi {0, 1, 2, 3, 4, 5, 6} --gpu 0 --random-transform --compute-val-loss --freeze-backbone --batch-size 32 --steps 1000 pascal|coco datasets/VOC2012|datasets/coco to start training. The init lr is 1e-3.
  • STEP2: python3 train.py --snapshot xxx.h5 --phi {0, 1, 2, 3, 4, 5, 6} --gpu 0 --random-transform --compute-val-loss --freeze-bn --batch-size 4 --steps 10000 pascal|coco datasets/VOC2012|datasets/coco to start training when val mAP can not increase during STEP1. The init lr is 1e-4 and decays to 1e-5 when val mAP keeps dropping down.

Evaluate

  1. PASCAL VOC

    • python3 eval/common.py to evaluate pascal model by specifying model path there.
    • The best evaluation results (score_threshold=0.01, mAP50) on VOC2007 test are:
    phi0
    mAP500.7896
    weights size17M

  2. MSCOCO

    • python3 eval/coco.py to evaluate coco model by specifying model path there.

Test

python3 inference.py to test your image by specifying image path and model path there.

image1 image2 image3