kinematic3d's People
Forkers
zhouleidcc chuanzhidong zfyong friendmine yanggui19891007 tianhaoyue devin-coder frizy-up pp-zhangyi shipingsong areslp qin-xs xxxvincent chengsq aipakchoi no-seaweed leo-wl zhangliliang penpen-thusilence lzw875 shawnrainman mostafa-mansour jianling233 ethanrylko jackweiwang pilotier neonhuang senwang98 yahyaghani zggg1p tuskaw wangluyabupt miniminke hideosneskinematic3d's Issues
Testing model on single image
Hi, thank you for sharing your work with us.
I'm curious is it possible to do inference on single image from my own data set using pretrained model? If yes, could you please provide us with necessary steps for testing?
the orientation decompose seems to fail according to the log ?
Hi, thanks for your great work . I train your project with mirror modification in network and loss. And I find something strange when seeing the training log. The axis and head accuracy is always around 0.50. It is the normal case that the acc of axis and head would converge to 1, right?
I wonder to know is this two item around 0.5 a strange phenomenon?
Thanks.
Train and test on Argoverse dataset
Hi,
Thank you for releasing the code!.
I would like to train and test the model on the Argoverse dataset.
Could you let me know the files/methods that I would have to change?
Thanks.
test_kitti_3d_forecast is not defined
Thanks for your great work!
i have a problem when i run train_pose.py
i change the conf.do_test and conf.do_ego from False to True.
But i find that test_kitti_3d_forecast is not defined.
Could you please give me some help?
some questions about runtime
i want to konw how you calculate the runtime.I see the M3D-RPN is 0.16s/im and you use the 1080ti.
And i use the 1080 i got it 0.165s/im,which is almost the same with yours.
But when i test the kinematic runtime i got unreasonable results.
When i use 3090 testing kinematic runtime. the test_kitti_3d_kalman_boxes only costs 43s,which is 43/3769=0.011s/im.I think this is unreasonable.But i find the extract_kalman_boxes costs 499s. So i want to konw extract_kalman_boxes is doing what?
Could you please give me some advice?
Inference in KITTI test
Thanks for your great work and I've learned a lot from this paper.
I have one question about your paper. You mentioned that "Inference uses 4 frames in KITTI". To the best of my knowledge, KITTI test set only provide 7,518 testing images while it does not contain the extra 4 temporally adjacent frames. So how do you perform inference on KITTI test?
Thanks and looking forward to your reply!
TypeError: zip argument #1 must support iteration
Hi,
Thanks for the great work! I meet the following error when I installing this repo, which I met the same error in M3DRPN framework. I guess it is a version problem again. I followed installation guide step by step, and I did everything except set CUDA to 10.2 (I have 10.0 instead "conda install pytorch==1.2.0 torchvision==0.4.0 cudatoolkit=10.0 -c pytorch").
Would the different version of CUDA on the same pytorch be the problem?
(kinematic3d) robot1@name-x99:~/Shukai/kinematic3d$ python scripts/train_rpn_3d.py --config=kitti_3d_warmup
Preloading imdb.
weighted respectively as 1.05 and 0.00
Found 3534 foreground and 178 empty images
Labels not used.. ['DontCare', 'Truck', 'Tram', 'Misc', 'Person_sitting']
conf: {
model: densenet121_3d_dilate_decomp_alpha
solver_type: sgd
lr: 0.004
momentum: 0.9
weight_decay: 0.0005
max_iter: 80000
snapshot_iter: 20000
display: 250
do_test: True
fast_eval: True
lr_policy: poly
lr_steps: None
lr_target: 4e-08
rng_seed: 5
cuda_seed: 8
image_means: [0.485, 0.456, 0.406]
image_stds: [0.229, 0.224, 0.225]
feat_stride: 16
has_3d: True
has_un: False
decomp_alpha: True
test_scale: 512
crop_size: [512, 1760]
mirror_prob: 0.5
distort_prob: -1
dataset_test: kitti_split1
datasets_train: [{'anno_fmt': 'kitti_det',
'im_ext': '.png',
'name': 'kitti_split1',
'scale': 1}]
use_3d_for_2d: True
percent_anc_h: [0.0625, 0.75]
min_gt_h: 32.0
max_gt_h: 384.0
min_gt_vis: 0.65
ilbls: ['Van', 'ignore']
lbls: ['Car', 'Pedestrian', 'Cyclist']
batch_size: 2
fg_image_ratio: 1.0
box_samples: 0.2
fg_fraction: 0.2
bg_thresh_lo: 0
bg_thresh_hi: 0.5
fg_thresh: 0.5
ign_thresh: 0.5
best_thresh: 0.35
nms_topN_pre: 3000
nms_topN_post: 40
nms_thres: 0.4
clip_boxes: False
test_protocol: kitti
test_db: kitti
test_min_h: 0
min_det_scales: [0, 0]
cluster_anchors: 0
even_anchors: 0
expand_anchors: 0
anchors: [[-0.5, -8.5, 15.5, 23.5, 51.969, 0.531,
1.713, 1.025, -0.799, 0.439, -1.085],
[-8.5, -8.5, 23.5, 23.5, 52.176, 1.618,
1.6, 3.811, -0.453, 0.031, -1.57],
[-16.5, -8.5, 31.5, 23.5, 48.334,
1.644, 1.529, 3.966, 0.673, -0.892,
-1.458],
[-2.528, -12.555, 17.528, 27.555,
44.781, 0.534, 1.771, 0.971, 0.093,
-0.072, -1.45],
[-12.555, -12.555, 27.555, 27.555,
44.704, 1.599, 1.569, 3.814, -0.187,
-0.378, -1.543],
[-22.583, -12.555, 37.583, 27.555,
43.492, 1.621, 1.536, 3.91, 0.719,
-0.806, -1.425],
[-5.069, -17.638, 20.069, 32.638,
34.666, 0.561, 1.752, 0.967, -0.384,
0.335, -1.78],
[-17.638, -17.638, 32.638, 32.638,
35.35, 1.567, 1.591, 3.81, -0.511,
-0.519, -1.558],
[-30.207, -17.638, 45.207, 32.638,
37.128, 1.602, 1.529, 3.904, 0.452,
-0.584, -1.504],
[-8.255, -24.01, 23.255, 39.01, 28.771,
0.613, 1.76, 0.98, 0.067, 0.212,
-1.555],
[-24.01, -24.01, 39.01, 39.01, 28.331,
1.543, 1.592, 3.66, -0.811, -0.312,
-1.508],
[-39.764, -24.01, 54.764, 39.01,
30.541, 1.626, 1.524, 3.908, 0.312,
-0.528, -1.419],
[-12.248, -31.996, 27.248, 46.996,
23.011, 0.606, 1.758, 0.996, 0.208,
0.151, -1.534],
[-31.996, -31.996, 46.996, 46.996,
22.948, 1.51, 1.599, 3.419, -1.076,
-0.441, -1.602],
[-51.744, -31.996, 66.744, 46.996,
25.0, 1.628, 1.527, 3.917, 0.334,
-0.414, -1.388],
[-17.253, -42.006, 32.253, 57.006,
18.479, 0.601, 1.747, 1.007, 0.347,
0.075, -1.607],
[-42.006, -42.006, 57.006, 57.006,
18.815, 1.487, 1.599, 3.337, -0.862,
0.071, -1.603],
[-66.759, -42.006, 81.759, 57.006,
20.576, 1.623, 1.532, 3.942, 0.323,
-0.386, -1.409],
[-23.527, -54.553, 38.527, 69.553,
15.035, 0.625, 1.744, 0.917, 0.41,
0.201, -1.728],
[-54.553, -54.553, 69.553, 69.553,
15.346, 1.29, 1.659, 3.083, -0.275,
0.083, -1.57],
[-85.58, -54.553, 100.58, 69.553,
16.326, 1.613, 1.527, 3.934, 0.268,
-0.366, -1.397],
[-31.39, -70.281, 46.39, 85.281,
12.265, 0.631, 1.747, 0.954, 0.317,
0.205, -1.645],
[-70.281, -70.281, 85.281, 85.281,
11.878, 1.044, 1.67, 2.415, -0.211,
0.142, -1.502],
[-109.171, -70.281, 124.171, 85.281,
13.58, 1.621, 1.539, 3.961, 0.189,
-0.301, -1.36],
[-41.247, -89.994, 56.247, 104.994,
9.932, 0.61, 1.771, 0.934, 0.486,
0.161, -1.663],
[-89.994, -89.994, 104.994, 104.994,
8.949, 0.811, 1.766, 1.662, 0.08,
0.105, -1.694],
[-138.741, -89.994, 153.741, 104.994,
11.043, 1.61, 1.533, 3.899, 0.04,
-0.216, -1.44],
[-53.602, -114.704, 68.602, 129.704,
8.389, 0.604, 1.793, 0.95, 0.806,
0.091, -1.621],
[-114.704, -114.704, 129.704, 129.704,
8.071, 1.01, 1.751, 2.19, -0.076,
0.389, -1.88],
[-175.806, -114.704, 190.806, 129.704,
9.184, 1.606, 1.526, 3.869, -0.066,
-0.278, -1.416],
[-69.089, -145.677, 84.089, 160.677,
6.923, 0.627, 1.791, 0.96, 0.784,
0.049, -1.699],
[-145.677, -145.677, 160.677, 160.677,
6.784, 1.384, 1.615, 2.862, -1.035,
1.046, -1.688],
[-222.266, -145.677, 237.266, 160.677,
7.863, 1.617, 1.55, 3.948, -0.071,
-0.202, -1.466],
[-88.5, -184.5, 103.5, 199.5, 5.189,
0.66, 1.755, 0.841, 0.173, 0.246,
-2.019],
[-184.5, -184.5, 199.5, 199.5, 4.388,
0.743, 1.728, 1.381, 0.642, -0.031,
-1.714],
[-280.5, -184.5, 295.5, 199.5, 5.583,
1.583, 1.547, 3.862, -0.072, -0.166,
-1.432]]
bbox_means: [[0.0, 0.0, 0.053, -0.1, 0.006, -0.067,
0.187, 0.06, -0.021, 0.07, -0.019,
-0.1, 0.045]]
bbox_stds: [[0.145, 0.131, 0.257, 0.248, 0.17,
0.137, 3.765, 0.398, 0.106, 0.523,
1.929, 1.216, 0.615]]
anchor_scales: [32.0, 40.11, 50.276, 63.019, 78.991,
99.012, 124.106, 155.561, 194.989,
244.409, 306.354, 384.0]
anchor_ratios: [0.5, 1.0, 1.5]
hard_negatives: True
focal_loss: 0
cls_2d_lambda: 1
iou_2d_lambda: 1
bbox_2d_lambda: 0
bbox_3d_lambda: 1
bbox_axis_head_lambda: 0.35
bbox_un_lambda: 0
infer_2d_from_3d: False
score_thres: 0.75
bbox_un_dynamic: False
}
Traceback (most recent call last):
File "scripts/train_rpn_3d.py", line 190, in <module>
main(sys.argv[1:])
File "scripts/train_rpn_3d.py", line 128, in main
cls, prob, bbox_2d, bbox_3d, feat_size, rois, rois_3d, rois_3d_cen = rpn_net(images)
File "/home/robot1/anaconda3/envs/kinematic3d/lib/python3.6/site-packages/torch/nn/modules/module.py", line 547, in __call__
result = self.forward(*input, **kwargs)
File "/home/robot1/anaconda3/envs/kinematic3d/lib/python3.6/site-packages/torch/nn/parallel/data_parallel.py", line 153, in forward
return self.gather(outputs, self.output_device)
File "/home/robot1/anaconda3/envs/kinematic3d/lib/python3.6/site-packages/torch/nn/parallel/data_parallel.py", line 165, in gather
return gather(outputs, output_device, dim=self.dim)
File "/home/robot1/anaconda3/envs/kinematic3d/lib/python3.6/site-packages/torch/nn/parallel/scatter_gather.py", line 68, in gather
res = gather_map(outputs)
File "/home/robot1/anaconda3/envs/kinematic3d/lib/python3.6/site-packages/torch/nn/parallel/scatter_gather.py", line 63, in gather_map
return type(out)(map(gather_map, zip(*outputs)))
File "/home/robot1/anaconda3/envs/kinematic3d/lib/python3.6/site-packages/torch/nn/parallel/scatter_gather.py", line 63, in gather_map
return type(out)(map(gather_map, zip(*outputs)))
File "/home/robot1/anaconda3/envs/kinematic3d/lib/python3.6/site-packages/torch/nn/parallel/scatter_gather.py", line 63, in gather_map
return type(out)(map(gather_map, zip(*outputs)))
TypeError: zip argument #1 must support iteration
Any help is appreciated!
Best,
Shukai
How can you obtain the extra labels?
Hi, thanks for your great work.
I wonder to know that how can you obtain the extra labels in the provided raw_extra data?
In the extra KITTI raw data, you provided the label_2 and calib folders.
Thanks.
Uncertainty in inference usage?
In the paper, it says "At inference, we fuse the self-balancing confidence with the classification score as µ = c · ω".
Before seeing your code, I guess that µ is used as the confidence for NMS. But in your project, µ just as the final score for evaluation. Why don't you use it in the NMS? Would it cause the performance degradation ?
Thanks for your reply!
Sincerely.
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.