cardwing / Codes-for-Lane-Detection

Learning Lightweight Lane Detection CNNs by Self Attention Distillation (ICCV 2019)
MIT License
1.04k stars 333 forks source link

testing my own dataset #351

Open MittalHarsh27 opened 1 year ago

MittalHarsh27 commented 1 year ago

Hi, I am trying to test the algorithm on my own dataset of image size as 2448 * 880, but I am unable to generate result of f1 measure and recall scores based on the SCNN repository mentioned in SCNN Tensorflow. I have pre processed my data according to CULane as well and I didn't get any error while running tools/test_lane.py but I wasn't getting scores in the end.

Can you please help me with resolving this issue?

This is the output I get in the last

------------Configuration--------- anno_dir: ../../data/CULane/ detect_dir: ../../tools/output/vgg_SCNN_DULR_w9/ im_dir: ../../data/CULane/ list_im_file: ../../data/CULane/list/test.txt width_lane: 30 iou_threshold: 0.5 im_width: 2448 im_height: 880

Evaluating the results... tp: 0 fp: 480 fn: 222 finished process file precision: 0 recall: 0 Fmeasure: -nan