szq0214/DSOD

Issues

• Couldn't find any detections

  #33 opened 2 years ago by bei91
  1

• Question about pooling layer in your DSOD300

  #31 opened 2 years ago by zjuzuhe
  1

• No such File or Directory - Core Dumped

  #32 opened 2 years ago by zeehasham
  1

• Sixth_norm_mbox_priorbox step parameter is wrong

  #37 opened 2 years ago by Juggernaut93
  1

• Reproduced result is lower than the reported result

  #39 opened 2 years ago by youngwanLEE
  1

• Why the speed of DSOD is too slow in the process of training?

  #43 opened 2 years ago by defei-coder
  1

• Have you tried large input images (600 * 1000)?

  #28 opened 2 years ago by mqchen1993
  1

• about how to batch test many image files？

  #27 opened 2 years ago by hu5tao
  1

• the last pooling size mismatch

  #23 opened 2 years ago by zhangkaij
  1

• The training time

  #22 opened 2 years ago by Dido0o0
  2

• Have you tried DSOD512?

  #19 opened 2 years ago by youngwanLEE
  1

• Transition w/o Pooling Layer size mismatch

  #18 opened 2 years ago by qianyizhang
  1

• How to measure the inference time?

  #17 opened 2 years ago by youngwanLEE
  3

• The net that python script created is diffrent from any pre-trained model offered in this README.md?

  #14 opened 2 years ago by wangzhe0623
  11

• mean_values_.size() == 1 || mean_values_.size() == img_channels Specify either 1 mean_value or as many as channels: 1

  #12 opened 2 years ago by jacco09
  1

• Check failed: error == cudaSuccess (2 vs. 0) out of memory.How to deal with it?

  #11 opened 2 years ago by githubfeiwang
  6

• no advantage with VGG-ssd ?

  #10 opened 2 years ago by birdwcp
  4

• What is your trick to overcome the GPU memory constraints?

  #9 opened 2 years ago by gd2016229035
  5

• will release the trianing code?

  #8 opened 2 years ago by zimenglan-sysu-512
  3

• Couldn't find any detections

  #7 opened 2 years ago by miraclewkf
  12

• Can not detect the small object?

  #5 opened 2 years ago by 10183308
  1

• run detection test with one input image

  #4 opened 2 years ago by kaishijeng
  1

• cannot agree faster-rcnn cannot converge when training from scratch

  #2 opened 2 years ago by argman
  1

• How to get VOC12 annotations?

  #1 opened 2 years ago by siyiding1216
  1

• How long does model take when I have 20 000 images, 1 class with Quadro P4000

  #44 opened 5 years ago by duongdqq
  0

• NameError ：name ‘DSOD300_V3_Body’ is not defined. How to deal with it?

  #41 opened 6 years ago by Jngwl
  1

• Any Dockerfile?

  #42 opened 6 years ago by francesco-p
  0

• Why not public caffe train_val.prototxt?

  #40 opened 6 years ago by ujsyehao
  0

• training on new dataset

  #36 opened 6 years ago by radu-diaconescu13
  1

• grp-dsod pretrained models

  #38 opened 6 years ago by endymecy
  0

• How to prepare voc12 test lmdb

  #24 opened 7 years ago by lzx1413
  3

• Test on VOC2012

  #34 opened 7 years ago by DL-ljw
  2

• Did the DSOD model code have tensorflow version?

  #26 opened 7 years ago by Tzpchris
  1

• The result after relu activation function isn't used in grp-dsod.

  #35 opened 7 years ago by zjuzuhe
  1

• small batchsize has a lower mAP.

  #30 opened 7 years ago by DL-ljw
  3

• Video Detection

  #25 opened 7 years ago by skathi1991
  1

• About 1 Channel images

  #29 opened 7 years ago by didulle
  2

• DSOD Visualization Problem on Video Test

  #20 opened 7 years ago by Betterthinking
  0

• Did this model have MXNet version?

  #21 opened 7 years ago by xhzcyc
  0

• I find the DSOD300_pascal.py can't breakpoint，When I stop running, it can not find the latest snapshoot

  #15 opened 7 years ago by moyans
  4

• About the test GPU memory？

  #16 opened 7 years ago by moyans
  3

• what is the difference between DSOD300_pascal and DSOD300_pascal++?

  #13 opened 7 years ago by wellingtonyl
  2

• No module named model_libs

  #3 opened 7 years ago by miraclewkf
  0