search

vicoslab / mixed-segdec-net-comind2021

Official PyTorch implementation for "Mixed supervision for surface-defect detection: from weakly to fully supervised learning"
Other
290 stars 89 forks source link

About Training #11

Open Ann331 opened 3 years ago

Ann331 commented 3 years ago

Hi, thanks for this great work. But some problems arise when I want to get your perfect results. I can't get the correct ROC and PR curve by training only on the DAGM dataset. Is there anything wrong? Could you please give me some advice on training work? Thanks very much.

There are some details about my training log: "0603_2_FOLD_1 Epoch 50/50 ==> avg_loss_seg=0.00284, avg_loss_dec=0.00041, avg_loss=0.00006, correct=158/158, in 19.77s/epoch (fwd/bck in 16.59s/epoch) 0603_2_FOLD_1 VALIDATION || AUC=nan, and AP=nan, with best thr=1.000000 at f-measure=nan and FP=5, FN=0, TOTAL SAMPLES=575"

JakobBozic commented 3 years ago

Hi, i tried again to train a DAGM model, and ROC and AUC score calculations work ok for me. Did you check that you use the correct versions of scikit libraries? If you give me the parameters you used to get this error i can try to run training again.

Ann331 commented 3 years ago

Hi, i tried again to train a DAGM model, and ROC and AUC score calculations work ok for me. Did you check that you use the correct versions of scikit libraries? If you give me the parameters you used to get this error i can try to run training again.

There are my training parameters. Please help me to check if there is any error. Thank you very much. --------------------------------parameters---------------------------- python -u train_net.py \ --GPU=0 \ --DATASET=DAGM \ --RUN_NAME=0603_2 \ --DATASET_PATH=datasets/DAGM/ \ --RESULTS_PATH=./results \ --SAVE_IMAGES=True \ --DILATE=7 \ --EPOCHS=50 \ --LEARNING_RATE=1.0 \ --DELTA_CLS_LOSS=0.01 \ --BATCH_SIZE=1 \ --WEIGHTED_SEG_LOSS=True \ --WEIGHTED_SEG_LOSS_P=2 \ --WEIGHTED_SEG_LOSS_MAX=1 \ --DYN_BALANCED_LOSS=True \ --GRADIENT_ADJUSTMENT=True \ --FREQUENCY_SAMPLING=True \ --TRAIN_NUM=33 \ --NUM_SEGMENTED=45 \ --FOLD=1

JakobBozic commented 3 years ago

Hi, I tried your parameters and I could not replicate your issue.

Ann331 commented 3 years ago

Hi, I tried your parameters and I could not replicate your issue.

I will check my parameters again. Thanks for your trying.

sunboyan1024 commented 3 years ago

Hi!Could you tell me What is the AP value of SDD2 in your experimrnt?My AP is only 92.0%,and i don't know what is wrong?