search

wentaozhu / DeepLung

WACV18 paper "DeepLung: Deep 3D Dual Path Nets for Automated Pulmonary Nodule Detection and Classification"
Apache License 2.0
312 stars 143 forks source link

about detection result #22

Closed jinmnie closed 6 years ago

jinmnie commented 6 years ago

hi @wentaozhu, the nodule you detected is more than 130 thousand, but I only got 4 thousand. What's going wrong with the detection? The number worker is 0, The batch size is set to 4. Other parameters are the same with yours.

wentaozhu commented 6 years ago

How is the FROC performance difference?

jinmnie commented 6 years ago

5 percent lower

wentaozhu commented 6 years ago

I used bigger batch size. Try to use 16 or like that in my code and run it on 4 GPUs. The batch size does affect performance.