facebookresearch/MaskFormer

Issues

• ValueError

  #72 opened 2 years ago by LiuMingYy
  0

• Whether mask embedding can also be seen as a dynamic kernel parameter？

  #71 opened 2 years ago by Yzichen
  0

• Error in training

  #70 opened 2 years ago by ainneabid
  0

• How to train instance segmentation with COCO dataset?

  #64 opened 3 years ago by WuTao-CS
  6

• Unable to train the model

  #69 opened 2 years ago by kagawa588
  0

• Pascal_VOC12 is not working

  #55 opened 3 years ago by obitoSad
  2

• A few questions about the configuration files

  #68 opened 2 years ago by WBS-123
  0

• Question about Pixel decoder last Conv2d layer

  #67 opened 3 years ago by YaGami01
  0

• Question about different decoder head

  #65 opened 3 years ago by chenhaiwen
  1

• Extract only mask from the output

  #66 opened 3 years ago by SahilChachra
  0

• Problem running demo on COCO Panoptic

  #38 opened 3 years ago by marcomiglionico94
  1

• Question about transformer decoder

  #63 opened 3 years ago by CoinCheung
  2

• Inputs preprocessing

  #62 opened 3 years ago by FrancescoSaverioZuppichini
  2

• Get raw logits per pixel

  #60 opened 3 years ago by NielsRogge
  3

• errors occurred when running demo.py

  #37 opened 3 years ago by fjremnav
  4

• Query on the Matching Step

  #59 opened 3 years ago by sauradip
  0

• Result in log file.

  #56 opened 3 years ago by chhluo
  9

• Unable to open file (file signature not found)

  #58 opened 3 years ago by xiaoquan-code
  1

• About ignore value

  #57 opened 3 years ago by NickChang97
  1

• Attention mask in last Swin basic layer

  #52 opened 3 years ago by shkarupa-alex
  1

• the input size of Flops is 256x256?

  #50 opened 3 years ago by Sunting78
  4

• Question about general inference

  #54 opened 3 years ago by Trainingzy
  2

• The comparsion between per-pixel classification and masks classification.

  #53 opened 3 years ago by ydhongHIT
  1

• [Question] Corner case handling in the dataset mapper

  #48 opened 3 years ago by programath
  1

• How to compute the FLOPS reported in the paper

  #47 opened 3 years ago by YellowPig-zp
  2

• What are cityscapes panoptic results for Maskformer?

  #49 opened 3 years ago by lxtGH
  3

• Panoptic segmentation on Cityscapes

  #43 opened 3 years ago by ydhongHIT
  2

• semantic seg training Error

  #51 opened 3 years ago by Mollylulu
  3

• the model becomes very bad after I set the dice_weights to 0.

  #46 opened 3 years ago by SeasarStalion
  1

• Why coco model shouldn't pad image?

  #45 opened 3 years ago by clownrat6
  1

• cannot import name 'FakeQuantizeBase' from 'torch.quantization'

  #42 opened 3 years ago by dennis-yuanlei
  1

• question about ade20k benchmark learning setting

  #41 opened 3 years ago by Sunting78
  4

• Can MaskFormer support multi-node training?

  #44 opened 3 years ago by qianyuzqy
  2

• The error when training

  #39 opened 3 years ago by daixiaolei623
  4

• Question about parameter tuning

  #40 opened 3 years ago by lxa9867
  6

• Convergence speed of maskformer

  #31 opened 3 years ago by dingjiansw101
  1

• No instruction or code for coco panoptic segmentation training?

  #35 opened 3 years ago by kl2005ad
  3

• which version of pre-trained is used in paper?

  #36 opened 3 years ago by caopulan
  1

• ADE20k prepare pan_seg, two stuffs has the same color possibly result in miss calculation of results

  #32 opened 3 years ago by zbwxp
  3

• Mapillary Vistas per class metrics

  #29 opened 3 years ago by nlgranger
  1

• Query embedding not learnable?

  #33 opened 3 years ago by YellowPig-zp
  1

• Question about normalization (Mean/Std) value different from swin pretrained backbone

  #34 opened 3 years ago by MaureenZOU
  1

• A naive question

  #27 opened 3 years ago by cnnAndBn
  3

• Some questions about FLOPs

  #28 opened 3 years ago by liuyang-ict
  4

• Calculation of PQ for semantic segmentation dataset

  #30 opened 3 years ago by dingjiansw101
  2

• Model cannot converge when training coco panoptic task

  #23 opened 3 years ago by loliq
  1

• where is the tansformer code part?

  #24 opened 3 years ago by zimurui
  1

• Any instance segmentation results?

  #25 opened 3 years ago by chufengt
  1

• How to prepare ground truth for MaskFormer-fixed setting?

  #22 opened 3 years ago by LavenderLA
  1

• Failed to download https://dl.fbaipublicfiles.com/detectron2/ImageNetPretrained/torchvision/R-50.pkl

  #21 opened 3 years ago by xinguo2
  1