-
## 一言でいうと
物体検出を予測領域と実領域(ラベル)との組み合わせ問題として解く研究。CNNで得た特徴と空間位置を表すpositional encoding(PE)をTransformer(Encoder)に入力し、出力をTransformer(Decoder)に渡し物体検出用PEでクエリする。Decoderはクラス確率と物体(中央)位置を出力。ラベルとの適合から学習する
 now
-
File "D:\code\MonoDETR\lib\models\monodetr\matcher.py", line 83, in forward
cost_giou = -generalized_box_iou(box_cxcylrtb_to_xyxy(out_bbox), box_cxcylrtb_to_xyxy(tgt_bbox))
File "D:\code\Mon…
-
## ❓ How to train DETR with resnet18 backbone?
Describe what you want to do, including:
1. I'm trying to run training on my 2080ti with resnet18 backbone and getting an error
2. I started with de…
-
请问pytorch权重是paddle权重转换过来的嘛，是的话能提供下转换代码嘛，想转下o365的权重，thanks。
-
I am training DETR with a new backbone, but while validation I get all IOU metric: bbox result is zero. Is that normal or there is semothing wrong?
here are the results from the log file
"test_coco_…
-
Hello,
After every epoch rtdetr_pytorch makes a checkpoint file,
Which i want to disable , i want only the best and the last ckpt
I've tried to look for in configs but i cannot find it anywhe…