Change stride number in detection_data_param causing error

[JingyaLiu]

While I change the stride number from 16 to 4 at detection_data_param and causing for following error, do you have any suggestions?

For train.prototxt: detection_data_param { root_folder: "/media/tensor-server/MedicalImage/LUNA16/" source: "../../../data/LUNA16/window_file_LUNA16_trainval.txt" batch_size: 1 short_size: 700 long_size: 1000 stride: 4 stride: 4 stride: 4 stride: 4 stride: 4 stride: 4 stride: 4 stride: 4 stride: 4 stride: 4 stride: 4 stride: 4 stride: 4 stride: 4 stride: 4

field_w: 8 field_w: 12 field_w: 10 field_w: 10 field_w: 30
field_w: 20 field_w: 40 field_w: 60  field_w: 50 field_w: 70
field_w: 90 field_w: 80 field_w: 90 field_w: 150 field_w: 120

field_h: 12 field_h: 8 field_h: 10 field_h: 30 field_h: 10
field_h: 20 field_h: 60 field_h: 40 field_h: 50 field_h:90
field_h: 70 field_h: 80  field_h: 150 field_h: 90 field_h:120

Error: I1029 22:39:25.876363 16399 layer_factory.hpp:77] Creating layer rpn_bbox_reshape_rpn_bbox_reshape_0_split I1029 22:39:25.876366 16399 net.cpp:84] Creating Layer rpn_bbox_reshape_rpn_bbox_reshape_0_split I1029 22:39:25.876369 16399 net.cpp:406] rpn_bbox_reshape_rpn_bbox_reshape_0_split <- rpn_bbox_reshape I1029 22:39:25.876374 16399 net.cpp:380] rpn_bbox_reshape_rpn_bbox_reshape_0_split -> rpn_bbox_reshape_rpn_bbox_reshape_0_split_0 I1029 22:39:25.876381 16399 net.cpp:380] rpn_bbox_reshape_rpn_bbox_reshape_0_split -> rpn_bbox_reshape_rpn_bbox_reshape_0_split_1 I1029 22:39:25.876386 16399 net.cpp:380] rpn_bbox_reshape_rpn_bbox_reshape_0_split -> rpn_bbox_reshape_rpn_bbox_reshape_0_split_2 I1029 22:39:25.876422 16399 net.cpp:122] Setting up rpn_bbox_reshape_rpn_bbox_reshape_0_split I1029 22:39:25.876426 16399 net.cpp:129] Top shape: 15 4 63 44 (166320) I1029 22:39:25.876428 16399 net.cpp:129] Top shape: 15 4 63 44 (166320) I1029 22:39:25.876431 16399 net.cpp:129] Top shape: 15 4 63 44 (166320) I1029 22:39:25.876433 16399 net.cpp:137] Memory required for data: 4604925552 I1029 22:39:25.876435 16399 layer_factory.hpp:77] Creating layer rpn_loss I1029 22:39:25.876451 16399 net.cpp:84] Creating Layer rpn_loss I1029 22:39:25.876453 16399 net.cpp:406] rpn_loss <- rpn_cls_reshape_rpn_cls_reshape_0_split_0 I1029 22:39:25.876457 16399 net.cpp:406] rpn_loss <- rpn_bbox_reshape_rpn_bbox_reshape_0_split_0 I1029 22:39:25.876461 16399 net.cpp:406] rpn_loss <- label_group_reshape_label_group_reshape_0_split_0 I1029 22:39:25.876466 16399 net.cpp:380] rpn_loss -> rpn_loss I1029 22:39:25.876536 16399 layer_factory.hpp:77] Creating layer rpn_loss F1029 22:39:25.876785 16399 detection_group_loss_layer.cpp:134] Check failed: height == bottom[2]->height() (63 vs. 250) Check failure stack trace: @ 0x7f170e73c5cd google::LogMessage::Fail() @ 0x7f170e73e433 google::LogMessage::SendToLog() @ 0x7f170e73c15b google::LogMessage::Flush() @ 0x7f170e73ee1e google::LogMessageFatal::~LogMessageFatal() @ 0x7f170ee5c510 caffe::DetectionGroupLossLayer<>::Reshape() @ 0x7f170ef584e2 caffe::Net<>::Init() @ 0x7f170ef5aa1e caffe::Net<>::Net() @ 0x7f170ef64c95 caffe::Solver<>::InitTrainNet() @ 0x7f170ef65225 caffe::Solver<>::Init() @ 0x7f170ef6551f caffe::Solver<>::Solver() @ 0x7f170ef85721 caffe::Creator_SGDSolver<>() @ 0x40dc12 train() @ 0x408b8b main @ 0x7f170ced2830 __libc_start_main @ 0x409569 _start @ (nil) (unknown) Aborted (core dumped)