search

ShaoqingRen / faster_rcnn

Faster R-CNN
Other
2.7k stars 1.22k forks source link

Some problems when i train VOC2007 database #193

Open EvanYyf opened 6 years ago

EvanYyf commented 6 years ago

opts: cache_name: 'fast_rcnn_VOC2007_ZF' conf: [1×1 struct] ignore_cache: 0 imdb: [1×1 struct] net_def_file: 'F:\yyf\faster_rcnn-master\faster_rcnn-master\models\fast_rcnn_prototxts\ZF\test.prototxt' net_file: 'F:\yyf\faster_rcnn-master\faster_rcnn-master\output\fast_rcnn_cachedir\fast_rcnn_VOC2007_ZF\voc_2007_trainval\final' roidb: [1×1 struct] suffix: ''

conf: batch_size: 128 bbox_thresh: 0.5000 bg_thresh_hi: 0.5000 bg_thresh_lo: 0.1000 fg_fraction: 0.2500 fg_thresh: 0.5000 image_means: [224×224×3 single] ims_per_batch: 2 max_size: 1000 rng_seed: 6 scales: 600 test_binary: 0 test_max_size: 1000 test_nms: 0.3000 test_scales: 600 use_flipped: 1 use_gpu: 1

faster_rcnn-master: test (voc_2007_test) 1/4952 time: 0.400s .....The middle part is omitted....... faster_rcnn-master: test (voc_2007_test) 1000/4952 time: 0.039s -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf

faster_rcnn-master: test (voc_2007_test) 1001/4952 time: 0.037s .....The middle part is omitted....... faster_rcnn-master: test (voc_2007_test) 3000/4952 time: 0.036s -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf

faster_rcnn-master: test (voc_2007_test) 3001/4952 time: 0.031s .....The middle part is omitted....... faster_rcnn-master: test (voc_2007_test) 4000/4952 time: 0.034s -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf

faster_rcnn-master: test (voc_2007_test) 4001/4952 time: 0.036s .....The middle part is omitted....... faster_rcnn-master: test (voc_2007_test) 4952/4952 time: 0.037s -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf -Inf

test all images in 182.725756 seconds. Cleared 0 solvers and 1 stand-alone nets aeroplane: pr: load: 639/4952 aeroplane: pr: load: 1275/4952 aeroplane: pr: load: 1870/4952 aeroplane: pr: load: 2495/4952 aeroplane: pr: load: 3092/4952 aeroplane: pr: load: 3691/4952 aeroplane: pr: load: 4279/4952 aeroplane: pr: load: 4892/4952 !!! aeroplane : 0.0000 0.0000 bicycle: pr: load: 638/4952 bicycle: pr: load: 1253/4952 bicycle: pr: load: 1850/4952 bicycle: pr: load: 2470/4952 bicycle: pr: load: 3074/4952 bicycle: pr: load: 3673/4952 bicycle: pr: load: 4262/4952 bicycle: pr: load: 4871/4952 !!! bicycle : 0.0000 0.0000 bird: pr: load: 640/4952 bird: pr: load: 1279/4952 bird: pr: load: 1877/4952 bird: pr: load: 2476/4952 bird: pr: load: 2991/4952 bird: pr: load: 3570/4952 bird: pr: load: 4168/4952 bird: pr: load: 4774/4952 !!! bird : 0.0000 0.0000 boat: pr: load: 639/4952 boat: pr: load: 1270/4952 boat: pr: load: 1868/4952 boat: pr: load: 2487/4952 boat: pr: load: 3081/4952 boat: pr: load: 3654/4952 boat: pr: load: 4237/4952 boat: pr: load: 4839/4952 !!! boat : 0.0000 0.0000 bottle: pr: load: 640/4952 bottle: pr: load: 1278/4952 bottle: pr: load: 1874/4952 bottle: pr: load: 2499/4952 bottle: pr: load: 3092/4952 bottle: pr: load: 3691/4952 bottle: pr: load: 4278/4952 bottle: pr: load: 4862/4952 !!! bottle : 0.0000 0.0000 bus: pr: load: 640/4952 bus: pr: load: 1276/4952 bus: pr: load: 1872/4952 bus: pr: load: 2493/4952 bus: pr: load: 3089/4952 bus: pr: load: 3679/4952 bus: pr: load: 4270/4952 bus: pr: load: 4874/4952 !!! bus : 0.0000 0.0000 car: pr: load: 640/4952 car: pr: load: 1277/4952 car: pr: load: 1872/4952 car: pr: load: 2493/4952 car: pr: load: 3091/4952 car: pr: load: 3677/4952 car: pr: load: 4269/4952 car: pr: load: 4874/4952 !!! car : 0.0000 0.0000 cat: pr: load: 640/4952 cat: pr: load: 1279/4952 cat: pr: load: 1876/4952 cat: pr: load: 2497/4952 cat: pr: load: 3092/4952 cat: pr: load: 3688/4952 cat: pr: load: 4274/4952 cat: pr: load: 4877/4952 !!! cat : 0.0000 0.0000 chair: pr: load: 640/4952 chair: pr: load: 1276/4952 chair: pr: load: 1874/4952 chair: pr: load: 2481/4952 chair: pr: load: 3077/4952 chair: pr: load: 3664/4952 chair: pr: load: 4249/4952 chair: pr: load: 4855/4952 !!! chair : 0.0000 0.0000 cow: pr: load: 640/4952 cow: pr: load: 1279/4952 cow: pr: load: 1873/4952 cow: pr: load: 2496/4952 cow: pr: load: 3091/4952 cow: pr: load: 3682/4952 cow: pr: load: 4271/4952 cow: pr: load: 4874/4952 !!! cow : 0.0000 0.0000 diningtable: pr: load: 637/4952 diningtable: pr: load: 1267/4952 diningtable: pr: load: 1865/4952 diningtable: pr: load: 2484/4952 diningtable: pr: load: 3081/4952 diningtable: pr: load: 3674/4952 diningtable: pr: load: 4263/4952 diningtable: pr: load: 4871/4952 !!! diningtable : 0.0000 0.0000 dog: pr: load: 638/4952 dog: pr: load: 1270/4952 dog: pr: load: 1868/4952 dog: pr: load: 2487/4952 dog: pr: load: 3085/4952 dog: pr: load: 3679/4952 dog: pr: load: 4270/4952 dog: pr: load: 4874/4952 !!! dog : 0.0000 0.0000 horse: pr: load: 640/4952 horse: pr: load: 1279/4952 horse: pr: load: 1878/4952 horse: pr: load: 2504/4952 horse: pr: load: 3096/4952 horse: pr: load: 3692/4952 horse: pr: load: 4279/4952 horse: pr: load: 4879/4952 !!! horse : 0.0000 0.0000 motorbike: pr: load: 640/4952 motorbike: pr: load: 1281/4952 motorbike: pr: load: 1880/4952 motorbike: pr: load: 2505/4952 motorbike: pr: load: 3098/4952 motorbike: pr: load: 3693/4952 motorbike: pr: load: 4279/4952 motorbike: pr: load: 4889/4952 !!! motorbike : 0.0000 0.0000 person: pr: load: 640/4952 person: pr: load: 1279/4952 person: pr: load: 1879/4952 person: pr: load: 2505/4952 person: pr: load: 3099/4952 person: pr: load: 3695/4952 person: pr: load: 4283/4952 person: pr: load: 4896/4952 !!! person : 0.0000 0.0000 pottedplant: pr: load: 640/4952 pottedplant: pr: load: 1281/4952 pottedplant: pr: load: 1883/4952 pottedplant: pr: load: 2506/4952 pottedplant: pr: load: 3103/4952 pottedplant: pr: load: 3699/4952 pottedplant: pr: load: 4284/4952 pottedplant: pr: load: 4898/4952 !!! pottedplant : 0.0000 0.0000 sheep: pr: load: 639/4952 sheep: pr: load: 1267/4952 sheep: pr: load: 1864/4952 sheep: pr: load: 2484/4952 sheep: pr: load: 3081/4952 sheep: pr: load: 3675/4952 sheep: pr: load: 4265/4952 sheep: pr: load: 4865/4952 !!! sheep : 0.0000 0.0000 sofa: pr: load: 640/4952 sofa: pr: load: 1280/4952 sofa: pr: load: 1878/4952 sofa: pr: load: 2505/4952 sofa: pr: load: 3092/4952 sofa: pr: load: 3691/4952 sofa: pr: load: 4268/4952 sofa: pr: load: 4870/4952 !!! sofa : 0.0000 0.0000 train: pr: load: 639/4952 train: pr: load: 1271/4952 train: pr: load: 1865/4952 train: pr: load: 2484/4952 train: pr: load: 3082/4952 train: pr: load: 3676/4952 train: pr: load: 4269/4952 train: pr: load: 4874/4952 !!! train : 0.0000 0.0000 tvmonitor: pr: load: 640/4952 tvmonitor: pr: load: 1281/4952 tvmonitor: pr: load: 1883/4952 tvmonitor: pr: load: 2506/4952 tvmonitor: pr: load: 3102/4952 tvmonitor: pr: load: 3699/4952 tvmonitor: pr: load: 4286/4952 tvmonitor: pr: load: 4899/4952 !!! tvmonitor : 0.0000 0.0000

Results:
     0
     0
     0
     0
     0
     0
     0
     0
     0
     0
     0
     0
     0
     0
     0
     0
     0
     0
     0
     0

     0

It seems that nothing had been computed ,how can i solve that?

EvanYyf commented 6 years ago

When the program goes to here,it starts to wrong.

Preparing training data...Done. Preparing validation data...Done.

------------------------- Iteration 2000 ------------------------- Training : error 0.698, loss (cls 2.66, reg NaN) Testing : error 0.692, loss (cls 2.59, reg NaN)

------------------------- Iteration 4000 ------------------------- Training : error 0.704, loss (cls 2.6, reg NaN) Testing : error 0.692, loss (cls 2.59, reg NaN)

------------------------- Iteration 6000 ------------------------- Training : error 0.686, loss (cls 2.57, reg NaN) Testing : error 0.692, loss (cls 2.58, reg NaN)

------------------------- Iteration 8000 ------------------------- Training : error 0.701, loss (cls 2.58, reg NaN) Testing : error 0.692, loss (cls 2.58, reg NaN)

------------------------- Iteration 10000 ------------------------- Training : error 0.688, loss (cls 2.58, reg NaN) Testing : error 0.692, loss (cls 2.59, reg NaN) Saved as F:\yyf\faster_rcnn-master\faster_rcnn-master\output\fast_rcnn_cachedir\fast_rcnn_VOC2007_ZF\voc_2007_trainval\iter_10000

------------------------- Iteration 12000 ------------------------- Training : error 0.702, loss (cls 2.6, reg NaN) Testing : error 0.692, loss (cls 2.58, reg NaN)

------------------------- Iteration 14000 ------------------------- Training : error 0.692, loss (cls 2.58, reg NaN) Testing : error 0.692, loss (cls 2.59, reg NaN)

------------------------- Iteration 16000 ------------------------- Training : error 0.686, loss (cls 2.56, reg NaN) Testing : error 0.692, loss (cls 2.58, reg NaN)

------------------------- Iteration 18000 ------------------------- Training : error 0.696, loss (cls 2.58, reg NaN) Testing : error 0.692, loss (cls 2.58, reg NaN)

------------------------- Iteration 20000 ------------------------- Training : error 0.692, loss (cls 2.58, reg NaN) Testing : error 0.692, loss (cls 2.58, reg NaN) Saved as F:\yyf\faster_rcnn-master\faster_rcnn-master\output\fast_rcnn_cachedir\fast_rcnn_VOC2007_ZF\voc_2007_trainval\iter_20000

------------------------- Iteration 22000 ------------------------- Training : error 0.69, loss (cls 2.57, reg NaN) Testing : error 0.692, loss (cls 2.58, reg NaN)

------------------------- Iteration 24000 ------------------------- Training : error 0.696, loss (cls 2.58, reg NaN) Testing : error 0.692, loss (cls 2.58, reg NaN)

------------------------- Iteration 26000 ------------------------- Training : error 0.696, loss (cls 2.58, reg NaN) Testing : error 0.692, loss (cls 2.59, reg NaN)

------------------------- Iteration 28000 ------------------------- Training : error 0.702, loss (cls 2.6, reg NaN) Testing : error 0.692, loss (cls 2.58, reg NaN)

------------------------- Iteration 30000 ------------------------- Training : error 0.689, loss (cls 2.56, reg NaN) Testing : error 0.682, loss (cls 2.55, reg NaN) Saved as F:\yyf\faster_rcnn-master\faster_rcnn-master\output\fast_rcnn_cachedir\fast_rcnn_VOC2007_ZF\voc_2007_trainval\iter_30000

------------------------- Iteration 32000 ------------------------- Training : error 0.677, loss (cls 2.52, reg NaN) Testing : error 0.679, loss (cls 2.5, reg NaN)

------------------------- Iteration 34000 ------------------------- Training : error 0.587, loss (cls 2.07, reg NaN) Testing : error 0.389, loss (cls 1.24, reg NaN)

------------------------- Iteration 36000 ------------------------- Training : error 0.325, loss (cls 1.05, reg NaN) Testing : error 0.266, loss (cls 0.867, reg NaN)

------------------------- Iteration 38000 ------------------------- Training : error 0.234, loss (cls 0.743, reg NaN) Testing : error 0.228, loss (cls 0.748, reg NaN) Saved as F:\yyf\faster_rcnn-master\faster_rcnn-master\output\fast_rcnn_cachedir\fast_rcnn_VOC2007_ZF\voc_2007_trainval\iter_40000 Saved as F:\yyf\faster_rcnn-master\faster_rcnn-master\output\fast_rcnn_cachedir\fast_rcnn_VOC2007_ZF\voc_2007_trainval\final