dvlab-research/PanopticFCN

Issues

• Inference on Cityscapes

  #45 opened 2 years ago by razvanbarbura
  2

• How to transfer this model to coreml?

  #50 opened 2 years ago by chaewonS
  0

• The result of cityscapes is strange

  #49 opened 2 years ago by canglangzhige
  1

• Inference on Cityscapes

  #48 opened 2 years ago by canglangzhige
  0

• 请问panopticfcn的网络支持转onnx吗？

  #47 opened 2 years ago by XXGYC
  1

• How to use the demo.py file？

  #46 opened 2 years ago by rongyiqing
  2

• Suspicious inference

  #43 opened 3 years ago by EnnioEvo
  3

• Reproducing Cityscapes results

  #21 opened 3 years ago by DdeGeus
  12

• Using the demo/demo.py

  #17 opened 3 years ago by mikkkeldp
  7

• No support for Cityscapes Dataset

  #44 opened 2 years ago by weiwwb
  1

• weakly supervised annotation generation

  #42 opened 3 years ago by xzxedu
  1

• How to get the visualization result?

  #32 opened 3 years ago by moddent
  1

• Setup project in Google Colab

  #41 opened 3 years ago by DianaUdisteanu
  0

• Loss became infinite or NaN when the cfg.MODEL.POSITION_HEAD.DEFORM is set to be True

  #33 opened 3 years ago by ShengkaiWu
  5

• 数据集问题

  #39 opened 3 years ago by kuozhijishou
  2

• Could you please share the download link of R-50-MSRA.pkl ?

  #40 opened 3 years ago by luoshuifeiyang
  2

• The " pretrained models" has the same URL

  #37 opened 3 years ago by qingchunnianshao
  1

• The purpose of embed_extractor in the thing/stuff_generator class

  #38 opened 3 years ago by Dwrety
  2

• reproducing results

  #31 opened 3 years ago by Yifeifr
  5

• I met a problem; the setting of max number of instances of the same category

  #35 opened 3 years ago by AntonotnaWang
  2

• Does the batch size matter?

  #28 opened 3 years ago by dspcad
  3

• train with cityscapes datasets

  #36 opened 3 years ago by DejaYang
  1

• About training on Cityscapes

  #30 opened 3 years ago by wudaxianzi
  5

• not enough values to unpack (expected 7, got 0)

  #34 opened 3 years ago by ZhaoNan1
  1

• How to train it with Cityscapes？

  #20 opened 3 years ago by ding-Chen
  4

• Data preparation and modification of config file for training with custom dataset

  #29 opened 3 years ago by mr-spielberg
  1

• About training on Mapillary dataset.

  #27 opened 3 years ago by HCShi
  7

• Compare with SOLOV2

  #25 opened 3 years ago by detectiveli
  3

• How can I test an image with a trained model?

  #26 opened 3 years ago by ding-Chen
  1

• SOLVER.STEPS are not modified in 3x config

  #19 opened 3 years ago by FreyWang
  1

• how to test image?

  #22 opened 3 years ago by szh91
  1

• About R101 corresponding configs and trained models

  #24 opened 3 years ago by ShihuaHuang95
  2

• Videos

  #23 opened 3 years ago by yaadhavraajagility
  0

• about batch size

  #18 opened 3 years ago by emjay73
  8

• Not sure how to use checkpoints

  #16 opened 3 years ago by mikkkeldp
  2

• Could you share the model on Cityscape as well?

  #4 opened 3 years ago by ChaoFang-TRI
  2

• What's the difference between PFCN and SOLOv2 on dynamic conv?

  #12 opened 3 years ago by lucasjinreal
  6

• runtime measurement

  #15 opened 3 years ago by emjay73
  4

• Is this model able to convert to tensorRT for faster acceleration?

  #1 opened 3 years ago by lucasjinreal
  5

• pretrained weigth name mismatch issue

  #14 opened 3 years ago by emjay73
  2

• some questions in code understanding

  #13 opened 3 years ago by Jing-W98
  0

• Can you share the code for predicting？

  #11 opened 3 years ago by x505367603
  4

• Demo script

  #10 opened 4 years ago by veer5551
  2

• Why it works well?

  #9 opened 4 years ago by CuberrChen
  3

• Question about kernel_fusion in head.py

  #7 opened 4 years ago by lucasjinreal
  5

• Training panoptic segmentation for a custom dataset in coco-panoptic format

  #8 opened 4 years ago by koushik-n
  4

• About the bias value for position head.

  #6 opened 4 years ago by GengDavid
  4

• difference between "FPN-Summed" and "Semantic FPN"

  #5 opened 4 years ago by valencebond
  2

• All 400 model leads to 512 size dowload link

  #3 opened 4 years ago by lucasjinreal
  4

• What's the difference between work of Panoptic Deeplab?

  #2 opened 4 years ago by lucasjinreal
  2