yjh0410 92faf28ffd rename YOLOvx to RTMDet-v1 2 vuotta sitten
..
README.md 92faf28ffd rename YOLOvx to RTMDet-v1 2 vuotta sitten
build.py 92faf28ffd rename YOLOvx to RTMDet-v1 2 vuotta sitten
loss.py 92faf28ffd rename YOLOvx to RTMDet-v1 2 vuotta sitten
matcher.py 92faf28ffd rename YOLOvx to RTMDet-v1 2 vuotta sitten
rtmdet_v1.py 92faf28ffd rename YOLOvx to RTMDet-v1 2 vuotta sitten
rtmdet_v1_backbone.py 92faf28ffd rename YOLOvx to RTMDet-v1 2 vuotta sitten
rtmdet_v1_basic.py 92faf28ffd rename YOLOvx to RTMDet-v1 2 vuotta sitten
rtmdet_v1_head.py 92faf28ffd rename YOLOvx to RTMDet-v1 2 vuotta sitten
rtmdet_v1_neck.py 92faf28ffd rename YOLOvx to RTMDet-v1 2 vuotta sitten
rtmdet_v1_pafpn.py 92faf28ffd rename YOLOvx to RTMDet-v1 2 vuotta sitten
rtmdet_v1_pred.py 92faf28ffd rename YOLOvx to RTMDet-v1 2 vuotta sitten

README.md

RTMDet-v1: My First Empirical Study of Real-Time General Object Detectors.

  • For training, we train my RTMDetv1 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 RTMDetv1-X with the setting of batch size=128.
Model Scale Batch APtest
0.5:0.95
APtest
0.5
APval
0.5:0.95
APval
0.5
FLOPs
(G)
Params
(M)
Weight
RTMDetv1-N 640 8xb16 9.1 2.4
RTMDetv1-T 640 8xb16 19.0 5.1
RTMDetv1-S 640 8xb16 43.6 62.6 43.3 62.6 33.6 9.0 ckpt
RTMDetv1-M 640 8xb16 48.3 67.0 48.1 66.9 87.4 23.6 ckpt
RTMDetv1-L 640 8xb16 50.2 68.6 50.0 68.4 176.6 47.6 ckpt
RTMDetv1-X 640