issues
search
HRNet
/
DEKR
This is an official implementation of our CVPR 2021 paper "Bottom-Up Human Pose Estimation Via Disentangled Keypoint Regression" (https://arxiv.org/abs/2104.02300)
MIT License
438
stars
76
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
长方形分辨率的图像训练效果很差
#61
wusaisa
opened
10 months ago
0
Is DEKR suitable for the task of single key point regression?
#60
CasonTsai
opened
10 months ago
0
How can I modify the code to fit my own dataset?
#59
wusaisa
opened
11 months ago
1
How can we obtain the bounding box for each person in the **'inference_demo'** ?
#58
zylofor
opened
1 year ago
0
Losses cannot be converged
#57
kuzuozhou97
opened
1 year ago
1
Where is the definition of eval?
#56
liudadan
opened
1 year ago
0
What is cfg? Where was it created?
#55
liudadan
opened
1 year ago
0
When is CocoKeypoints called
#54
liudadan
opened
1 year ago
0
export ONNX is fail
#53
xddlj
opened
1 year ago
0
Why can reading obj ['segmentation '] generate rle and rles? Rles is iterative, and rle is not iterative.
#52
liudadan
opened
1 year ago
0
Why do you re create the mask without using the label of the heatmap and the label of the paf as the mask
#51
liudadan
opened
1 year ago
0
Does the code call the CocoKeypoints (CocoDataset) class? I found that this class was not created in the code and was automatically called during training.
#50
liudadan
closed
1 year ago
1
none
#49
liudadan
closed
1 year ago
0
Can the code generate 19 dimension heatmaps?
#48
liudadan
opened
1 year ago
2
Unable to download the pre training model, please share the download link
#47
gjgjgjfff
closed
1 year ago
1
Visualization
#46
A7777-gp
opened
1 year ago
0
vision
#45
yuan243212790
opened
1 year ago
0
Confidence score for each joint
#44
anas-zafar
closed
2 years ago
2
Why not use mosaic, Have you tried it? 为什么不使用马赛克增强，有试过吗
#43
a1802lzw
closed
1 year ago
2
Queries regarding inference
#42
anas-zafar
opened
2 years ago
0
Take this code to pretrain the model to extract something similar to Market1501
#41
Dandelion-hi
opened
2 years ago
0
Training custom dataset
#40
wliebtzy
opened
2 years ago
0
Hi, can this network be used for hand joint point detection?
#39
mywordwin
closed
2 years ago
0
无法训练自己的数据
#38
wangduyang
opened
2 years ago
2
can it train on other dataset like posetrack?
#37
chenrxi
opened
2 years ago
0
Questions about the salient regions in figure 1
#36
Juzezhang
opened
2 years ago
0
Is there any comparison between adaptconv and deformable-conv?
#35
YTEP-ZHI
opened
2 years ago
0
KeyError: "There is no item named 'val2017/000000397133.jpg' in the archive"
#34
longpeace
opened
2 years ago
1
JOINT_COCO_LINK_1 and JOINT_COCO_LINK_2
#33
lcl180
opened
2 years ago
1
RutimeError:CUDA out of memory. when i training on COCO train2017 dataset
#32
houshuaishuai
opened
2 years ago
2
about offset loss, difficult to convergence?
#31
jianzfb
opened
2 years ago
3
When I execute the first command in python tools/valid.py \ --cfg experiments/coco/rescore_coco.yaml \ TEST.MODEL_FILE model/pose_coco/pose_dekr_hrnetw32.pth, I can't find '../data/rescore_data/rescore_dataset_train_coco_kpt' in it, and an error will be reported after the execution of more than 110000 pictures. I think it's because '../data/rescore_data/rescore_dataset_train_coco_kpt' needs to be created by myself, so I created such a file and modified the path in 'score/data/rescore_data/recore_dataset_train_coco_kpt', but an error is still reported. I hope you can help me solve it. In addition, I have another question. I have changed it three times. Each time I need to run 110000 pictures first, so it takes a lot of time. Is there any way to make him not need to run these 110000 pictures again?
#30
LiPan123456789
opened
2 years ago
2
When I was training the scoring net, model/pose_coco/pose_dekr_hrnetw32_coco.pth in the command could not be found. It was a file, so I changed it to ‘output/coco_kpt/hrnet_dekr/w32_4x_reg03_bs10_512_adam_lr1e-3_coco_x140/final_state0.pth.tar’. is this command correct, because he reported an error that the picture size is inconsistent? I can't find how to change the picture size. Please help me solve it~
#29
LiPan123456789
closed
2 years ago
4
When I was running the training of crowdpose dataset, there were 'nan' and 'NaN or Inf found in input tensor' problems. I checked the data. This problem is caused by the disappearance or explosion of the gradient. After I adjusted the learning rate, this problem still can not be solved. Please help me.
#28
LiPan123456789
opened
2 years ago
1
In the using inference demo, where is the video file/ multi_ people. mp4?
#27
LiPan123456789
closed
2 years ago
2
How to run DEKR in real-time
#26
ViktorTruongTuan
opened
2 years ago
1
how's the model performance if train on wholebody dataset？
#25
lucasjinreal
opened
2 years ago
1
Fix model path error
#24
jenhaoyang
closed
2 years ago
0
RuntimeError:cuDNN error:CUDNN_STATUS_NOT_INITIALIZED
#23
sysuzgg
opened
2 years ago
9
how can i vis the heatmap_avg
#22
CUMTweiwenxuan
opened
2 years ago
1
About AdaptBlock
#21
captainfffsama
closed
2 years ago
1
ONNX file generation fails
#20
vp211991
opened
3 years ago
2
Replicate experiments for W48
#19
bmartacho
opened
3 years ago
3
installation issue: ncclSystemError: System call (socket, malloc, munmap, etc) failed.
#18
looninho
closed
2 years ago
3
How to test on a single image？ I‘m confused to use the offsets.
#17
Oranging1
opened
3 years ago
0
Exhausted GPU memory with batch size of 1
#16
wmcnally
closed
3 years ago
2
inference_demo.py csv_header little bug
#15
yyccli
closed
3 years ago
1
CrowdPose Api bug fixed
#14
yyccli
closed
3 years ago
1
set_epoch for DistributedSampler
#13
ArchNew
closed
3 years ago
4
about training loss
#12
shanyun123456
closed
3 years ago
7
Next