Open leoluopy opened 4 years ago
what are the main differences between v2 and v1 ? and is the pretrained model afforded ? are they training scales or some thing else ?
v1 - refer to the paper for details v2 - the detection scale is 10-320 (vs 10-560 in v1), the number of layers is 20, the backbone is modified for faster inference. Refer to ./symbol_farm/symbol_structures.xlsx for details.
what are the main differences between v2 and v1 ? and is the pretrained model afforded ? are they training scales or some thing else ?