search

AlibabaResearch / efficientteacher

A Supervised and Semi-Supervised Object Detection Library for YOLO Series
GNU General Public License v3.0
831 stars 151 forks source link

gt_num非常低，怎么解决啊 #151

Open mingtianyidingtian opened 2 weeks ago

mingtianyidingtian commented 2 weeks ago 
 Epoch   gpu_mem    labels  img_size       box       obj       cls      loss    ss_box    ss_obj    ss_cls        tp    fp_cls    fp_loc   pse_num    gt_num
 12/19     13.5G      2013       640   0.03916   0.04745   0.00286     2.806   0.01344   0.01261 0.0003967  0.009873         0    0.3827     8.491   0.09531: 100%|██████████████████████████| 120/120 [09:18<00:00,  4.66s/it] 
           Class     Images     Labels          P          R     mAP@.5 mAP@.5:.95: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████| 2/2 [00:01<00:00,  1.83it/s] 
             all         90       1673      0.799      0.778      0.822      0.509
           Class     Images     Labels          P          R     mAP@.5 mAP@.5:.95: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████| 2/2 [00:01<00:00,  1.61it/s] 
             all         90       1673      0.801      0.803      0.834      0.542

 Epoch   gpu_mem    labels  img_size       box       obj       cls      loss    ss_box    ss_obj    ss_cls        tp    fp_cls    fp_loc   pse_num    gt_num
 13/19     13.5G      1265       640   0.03936   0.04876  0.002523     2.846   0.01218  0.008742 0.0002953  0.004488         0    0.3565     6.111   0.04766: 100%|██████████████████████████| 120/120 [08:45<00:00,  4.38s/it] 
           Class     Images     Labels          P          R     mAP@.5 mAP@.5:.95: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████| 2/2 [00:01<00:00,  1.89it/s]
             all         90       1673      0.803      0.782      0.826       0.51
           Class     Images     Labels          P          R     mAP@.5 mAP@.5:.95: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████| 2/2 [00:01<00:00,  1.61it/s]
             all         90       1673      0.794      0.806      0.835      0.541
mingtianyidingtian commented 2 weeks ago

SSOD: train_domain: True nms_conf_thres: 0.1 nms_iou_thres: 0.65 teacher_loss_weight: 1.0 cls_loss_weight: 0.3 box_loss_weight: 0.05 obj_loss_weight: 0.7 loss_type: 'ComputeStudentMatchLoss' ignore_thres_low: 0.3 ignore_thres_high: 0.8 uncertain_aug: True use_ota: False multi_label: False ignore_obj: False pseudo_label_with_obj: True pseudo_label_with_bbox: True pseudo_label_with_cls: False with_da_loss: True da_loss_weights: 0.01 epoch_adaptor: True resample_high_percent: 0.25 resample_low_percent: 0.99 ema_rate: 0.999 cosine_ema: True imitate_teacher: False

dynamic_thres: True

ssod_hyp: with_gt: False mosaic: 1.0 cutout: 0.5 autoaugment: 0.5 scale: 0.8 degrees: 0.0 shear: 0.0 project: 'yolov5_ssod' adam: False epochs: 20

weights: 'D:\project\efficientteacher-main\scripts\mula_convertor\efficent-spore.pt'

weights: 'D:/project/efficientteacher-main/yolov5_ssod/exp8/weights/best.pt' prune_finetune: False linear_lr: True find_unused_parameters: True

hyp: lr0: 0.001 hsv_h: 0.015 hsv_s: 0.7 hsv_v: 0.4 lrf: 1.0 scale: 0.9 burn_epochs: 0 no_aug_epochs: 0

mixup: 0.1

warmup_epochs: 0