YOLOv3_tf

Try to implement YOLOv3 with TensorFlow

1. for now, only train detector layer(50 epoch, batch size: 8)
2. froze feature extractor('darknet53.conv.74')
3. TODO: finetune end2end

traindata: VOC2007 trainval & VOC2012 trainval

1. TensorFlow 1.*
2. easydict
3. PIL
4. numpy
5. matplotlib

Name		Name	Last commit message	Last commit date
Latest commit History 12 Commits
ckpt		ckpt
font		font
image		image
.gitignore		.gitignore
README.md		README.md
config.py		config.py
create_tfrecords.py		create_tfrecords.py
darknet53_trainable.py		darknet53_trainable.py
data_pipeline.py		data_pipeline.py
draw_boxes.py		draw_boxes.py
load_weights.c		load_weights.c
predict_net.py		predict_net.py
prediction.jpg		prediction.jpg
save_npz.py		save_npz.py
test2007.txt		test2007.txt
test_darknet53.py		test_darknet53.py
test_mask.py		test_mask.py
test_movingavg.py		test_movingavg.py
test_upsampling.py		test_upsampling.py
test_yolo_head.py		test_yolo_head.py
test_yolov3.py		test_yolov3.py
train_net.py		train_net.py
yolo_layer.py		yolo_layer.py
yolo_top.py		yolo_top.py
yolov3-voc.cfg		yolov3-voc.cfg

anguoyang/YOLOv3_tf