yjh0410 fca944e333 update		2 lat temu
..
README.md	6f4f753562 keep training YOLOX-L	2 lat temu
build.py	2eb02320fc release RTCDet-v1 & rename RTMDet to RTCDet	2 lat temu
loss.py	fca944e333 update	2 lat temu
matcher.py	8d590a1158 modify RTCDet-v2	2 lat temu
rtcdet_v2.py	5754e59167 debug RTCDet-v2	2 lat temu
rtcdet_v2_backbone.py	fd8e9e5e0c update ELANNet-v2	2 lat temu
rtcdet_v2_basic.py	fd8e9e5e0c update ELANNet-v2	2 lat temu
rtcdet_v2_head.py	fd8e9e5e0c update ELANNet-v2	2 lat temu
rtcdet_v2_neck.py	fd8e9e5e0c update ELANNet-v2	2 lat temu
rtcdet_v2_pafpn.py	fd8e9e5e0c update ELANNet-v2	2 lat temu
rtcdet_v2_pred.py	767ac95ea3 add aux loss for RTCDet-v2	2 lat temu

RTCDet-v2: My Second Empirical Study of Real-Time Convolutional Object Detectors.

For training, we train my RTCDetv2 series series with 300 epochs on COCO.
For data augmentation, we use the large scale jitter (LSJ), Mosaic augmentation and Mixup augmentation, following the setting of YOLOX, but we remove the rotation transformation which is used in YOLOX's strong augmentation.
For optimizer, we use AdamW with weight decay 0.05 and base per image lr 0.001 / 64.
For learning rate scheduler, we use linear decay scheduler.
Due to my limited computing resources, I can not train RTCDetv2-X with the setting of batch size=128.