issues
search
microsoft
/
scene_graph_benchmark
image scene graph generation benchmark
MIT License
380
stars
86
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
how to set DATASETS FACTORY_TEST(when extract features from my own directory of images)
#52
BigHyf
opened
2 years ago
1
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
BigHyf
closed
2 years ago
0
How to generate the predicted object attributes and relations labels together?
#50
yangapku
opened
2 years ago
0
how to export to pb?
#49
KrystalCWT
opened
2 years ago
1
Why does the program freeze when running with multiple GPUs?
#48
tao123322
opened
2 years ago
0
TypeError: run_test() got an unexpected keyword argument'model_name
#47
tao123322
opened
2 years ago
5
KeyError: 'gt_labels' during training
#46
yizhe-ang
opened
2 years ago
3
missing model weight when training on visualgenome
#45
entalent
opened
2 years ago
6
made path in tutorial relative rather than absolute
#44
alliedel
opened
2 years ago
0
fail to setup
#43
KrystalCWT
closed
2 years ago
1
About Vinvl R50-C4 Model
#42
LittleDonkey1203
opened
2 years ago
0
Execute setup.py with the problem link.exe failed with code 1181 in windows platform
#41
destructive-observer
opened
2 years ago
0
Is that a bug for extracting visual features?
#40
youngfly11
opened
2 years ago
0
KeyError: 'box_features' --- extracting features from own image tsv files
#39
sangwoo0123
opened
2 years ago
3
About attribute and object lable for the pointed or designated bounding box
#38
alice-cool
opened
2 years ago
0
Cannot run VinVL feature extraction command
#37
lukerm
opened
2 years ago
9
Pre-extracted Image Features: what OD model is used?
#36
ahnjaewoo
closed
2 years ago
1
Index error when accessing box features for RelDN model
#35
aleSuglia
opened
2 years ago
0
ModelZoo contains broken links
#34
aleSuglia
closed
2 years ago
2
config_file for train_sg_net.py
#33
aaronhd
opened
2 years ago
1
cannot download pretrained model
#32
hapticmusic
closed
2 years ago
0
Issue when attempting to generate image features
#31
Frederick369
opened
2 years ago
1
Question about image ids
#30
uehara-mech
opened
3 years ago
5
How to generate train.labelmap.tsv ?
#29
Edwardmark
opened
3 years ago
7
Guide to run train_net.py
#28
Edwardmark
opened
3 years ago
2
Several issues in extracting VinVL Feature extraction
#27
he159ok
closed
3 years ago
2
Torch17
#26
hanxiaotian
closed
3 years ago
1
Could you please show me an easy way to generate features for single images?
#25
Edwardmark
closed
2 years ago
4
When I extract image features with VinVL, the AssertionError occur
#24
PengPeixi
opened
3 years ago
8
Res152c4 on 4 datasets seems not right
#23
helson73
opened
3 years ago
2
VinVL can model the relation prediction?
#22
alice-cool
closed
2 years ago
5
the labelmap only provide 50 objects? Could it more bigger Or we only from scratch to train the project object detection?
#21
alice-cool
opened
3 years ago
6
MIssing links to pre-trained model checkpoints
#20
coldmanck
closed
3 years ago
1
predcls ValueError: object labelmap is required, but was not provided
#19
alice-cool
opened
3 years ago
11
[Errno 2] No such file or directory: './datasets/openimages_v5c/vrd/ji_vrd_labelmap.json'
#18
alice-cool
closed
3 years ago
0
Inference on COCO/NoCaps dataset
#17
opetliak
opened
3 years ago
2
How to extract relation features
#16
moneyQyh
opened
3 years ago
6
Image file in Vinvl example script
#15
jerry0815
closed
3 years ago
1
Download vinvl pre-training model
#14
yanan1989
closed
3 years ago
5
RuntimeError: CUDA error: invalid device function
#13
mohaoran93
closed
3 years ago
3
How to extract the scene graph?
#12
JXT218
closed
3 years ago
5
Extracting features for a directory of images
#11
tejas1995
closed
3 years ago
1
demo_image.py not working
#10
JXT218
closed
3 years ago
3
demo yaml
#9
JXT218
closed
3 years ago
1
Slow feature extraction compared to bottom-up-attention
#8
vinson2233
closed
3 years ago
7
Correct way to extract image features with VinVL
#7
stopmosk
opened
3 years ago
14
where can I find the dataset yaml in the config file?
#6
runzeer
closed
3 years ago
7
Broken links to VinVL model and associated labelmaps
#5
stopmosk
closed
3 years ago
2
git clone https://github.com/hanxiaotian/scene_graph_benchmark.git
#4
mohaoran93
closed
3 years ago
2
Develop
#3
hanxiaotian
closed
3 years ago
1
Previous
Next