yjh0410 9065e15f1a add HeadConv in RTCDet PaFPN		2 سال پیش
..
README.md	9065e15f1a add HeadConv in RTCDet PaFPN	2 سال پیش
build.py	73011ee555 debug YOLOX-style Transform with Rotation	2 سال پیش
loss.py	efa0bed448 debug YOLOX-style Transform with Rotation	2 سال پیش
matcher.py	2727b6adb1 debug trainer	2 سال پیش
yolox.py	73011ee555 debug YOLOX-style Transform with Rotation	2 سال پیش
yolox_backbone.py	8f4509bd61 modify YOLOX	2 سال پیش
yolox_basic.py	3246f3efdd update	2 سال پیش
yolox_head.py	311a9b89b7 fix a unknown bug in YOLOX	2 سال پیش
yolox_neck.py	8f4509bd61 modify YOLOX	2 سال پیش
yolox_pafpn.py	3246f3efdd update	2 سال پیش

YOLOX:

For training, we train YOLOX series with 300 epochs on COCO.
For data augmentation, we use the large scale jitter (LSJ), Mosaic augmentation and Mixup augmentation.
For optimizer, we use SGD with weight decay 0.0005 and base per image lr 0.01 / 64,.
For learning rate scheduler, we use Cosine decay scheduler.
The reason for the low performance of my reproduced YOLOX-L has not been found out yet.

Model	Backbone	Batch	Scale	AP^val 0.5:0.95	AP^val 0.5	FLOPs ^(G)	Params ^(M)	Weight
YOLOX-S	CSPDarkNet-S	8xb8	640	40.1	60.3	26.8	8.9	ckpt
YOLOX-M	CSPDarkNet-M	8xb8	640	46.2	66.0	74.3	25.4	ckpt
YOLOX-L	CSPDarkNet-L	8xb8	640	48.7	68.0	155.4	54.2	ckpt
YOLOX-X	CSPDarkNet-X	8xb8	640