microsoft/scene_graph_benchmark

Issues

• Emergency, the pretrained model download links are lost

  #97 opened 2 years ago by yxding95
  7

• broken links

  #103 opened 2 months ago by niloofarAzari
  0

• problem of loading dataset, please help me!

  #102 opened 7 months ago by zetaodu
  0

• Alternate location for Vinvl ckpt and labelmap files

  #101 opened a year ago by sarvghotra
  0

• No SAS

  #100 opened a year ago by Lxy811
  0

• how to download this file: visualgenome/VG-SGG-dicts-danfeiX-clipped.json

  #63 opened 3 years ago by kingdomrush2
  2

• after training on own datasets getting zeros on result

  #91 opened 2 years ago by tanuj1221
  1

• download links of pre-trained models are invalid

  #94 opened 2 years ago by emptybird
  2

• I got nan losses on results

  #95 opened 2 years ago by LIUTAOGE
  2

• Invalid Link

  #99 opened a year ago by shed-e
  0

• How to add masks for train

  #98 opened 2 years ago by baidingyuan
  0

• [VinVL] Support for ONNX

  #96 opened 2 years ago by aleSuglia
  1

• How can I train without any problems?

  #64 opened 3 years ago by sunn-u
  2

• Runtime error occurred in Image Ids: CUDA error: device-side assert triggered

  #83 opened 3 years ago by albertmundu
  7

• Is the pre-trained Resnet-50 object detection model available?

  #93 opened 2 years ago by VSJMilewski
  0

• Extremely different evaluation results on Visual Genome using RelDN

  #74 opened 3 years ago by smichniak
  6

• Missing pretained model of X152fpn

  #90 opened 2 years ago by Ceazy-Gentleman
  0

• Updating detector from maskrcnn to detectron

  #78 opened 3 years ago by DavidHuji
  2

• CUDA error: device-side assert triggered

  #58 opened 3 years ago by hszhoushen
  1

• Problem with training TSVYamlDataset

  #89 opened 2 years ago by sho8circ8
  0

• Upgrade to PyTorch 1.11

  #87 opened 3 years ago by dreamflasher
  3

• using tsv_demo.py to generate some tsv files using my own images

  #53 opened 3 years ago by BigHyf
  1

• Runtime error occurred when extracting features for VinVL (running train_sg_net.py).

  #73 opened 3 years ago by KatieGou
  0

• Setup failed

  #56 opened 3 years ago by MeiqiGuo
  2

• Docker image broken

  #69 opened 3 years ago by dreamflasher
  6

• error from :assert l_batch == 1

  #86 opened 3 years ago by catfish132
  1

• when I was running motif net with 'R152FPN_motif_oi.yaml', an attribute error happened.

  #85 opened 3 years ago by catfish132
  0

• How to use pretrained neural motif to generate scene graph in my own image?

  #84 opened 3 years ago by Thinking-more
  1

• How to extract relation scores

  #82 opened 3 years ago by ForawardStar
  0

• Can I test with VinVL features under SGG?

  #81 opened 3 years ago by Pter61
  0

• Couldn't load custom C++ ops.

  #60 opened 3 years ago by QC-LY
  5

• Where to find maskrcnn-benchmark-1 ?

  #80 opened 3 years ago by dreamflasher
  0

• Please provide all config files: vinvl_x152c4.yaml test_vgoi6_clipped.yaml etc.

  #79 opened 3 years ago by dreamflasher
  1

• how to set DATASETS FACTORY_TEST(when extract features from my own directory of images)

  #52 opened 3 years ago by BigHyf
  1

• AssertionError?

  #77 opened 3 years ago by via815via
  2

• the labels of my own datasets don't have "rect" (xyxy mode)

  #76 opened 3 years ago by via815via
  0

• Results are all zeros

  #72 opened 3 years ago by DavidHuji
  0

• AssertionError: Override list has odd length:

  #71 opened 3 years ago by gzSYSUmrlian
  0

• Conflicting version pytorch and torchvision

  #68 opened 3 years ago by rafaelpadilla
  5

• how to download this file: visualgenome/VG-SGG-dicts-vgoi6-clipped.json

  #70 opened 3 years ago by mattsetzler
  1

• Is there a way to control the number of the predicted bounding boxes?

  #67 opened 3 years ago by JingbiaoMei
  0

• KeyError: 533 Issue with extracting features for vinvl

  #66 opened 3 years ago by shizhediao
  0

• test_sg_net for oscar image captioning fails because of features

  #65 opened 3 years ago by onuraydemir7
  0

• CUDA error: device-side assert triggered

  #61 opened 3 years ago by DavidHuji
  1

• xxxx

  #62 opened 3 years ago by korcys
  0

• How to download the dataset for GQA?

  #59 opened 3 years ago by uehara-mech
  0

• Object detection for predcls

  #57 opened 3 years ago by SenJia
  0

• can you add same GCN methods in recent CVPR

  #55 opened 3 years ago by Rh-Dang
  0

• con't not download openimages_v5c

  #54 opened 3 years ago by ihaeyong
  0

• Hi, I found a way to circumvent using tsv files by modifying `scene_graph_benchmark/tools/demo/demo_image.py`, and now I only need `jpg` image dataset, VinVl yaml configuration file and model weight file. The predictions are saved in dictionary and are stored in `pth` format. I ran it on Google Colab and it generates predictions at a rate about 2s/image. I hope this helps.

  #51 opened 3 years ago by BigHyf
  0