RELEARNING
IMAGE RECOGNITION
VOCAB
ULARY
IN FIVE
LAYERS
LOWER LAYERS
DEFINE UPPER ONES
NO CODE
THIS TIME
0
MODEL
LOSS
BACKBONE
HEAD
1
DETECT
SEGMENT
DEPTH
OBB
2
DINOv3
YOLO26
SAM 3
3
OCCLUSION
4
TOOLS & LICENSE
KERNEL / EMBEDDING / THRESHOLD
ALREADY MEASURED