swords123 / ida-3d Goto Github PK
View Code? Open in Web Editor NEWIDA-3D: Instance-Depth-Aware 3D Object Detection from Stereo Vision for Autonomous Driving, in CVPR2020
License: MIT License
IDA-3D: Instance-Depth-Aware 3D Object Detection from Stereo Vision for Autonomous Driving, in CVPR2020
License: MIT License
您好,请问一下论文中提到的一秒12帧是全部的运行时间还是只包括IDA模块?
I followed instructions to run the code on singularity. But when I run test_net.py. There is an error occurred. Does my local CUDA version(11.0) cause this issue?
(tdrcnn) Singularity> python test_net.py --config-file self_exp/exp_1.yal --ckpt self_exp/exp_1/model_0080000.pth
Traceback (most recent call last):
File "test_net.py", line 15, in
from maskrcnn_benchmark.engine.inference import inference
File "/home/usr1/Desktop/project/IDA-3D/maskrcnn_benchmark/engine/inference.py", line 10, in
from maskrcnn_benchmark.data.datasets.evaluation import evaluate
File "/home/usr1/Desktop/project/IDA-3D/maskrcnn_benchmark/data/datasets/evaluation/init.py", line 3, in
from .kitti import kitti_evaluation
File "/home/usr1/Desktop/project/IDA-3D/maskrcnn_benchmark/data/datasets/evaluation/kitti/init.py", line 3, in
from .kitti_eval import do_kitti_evaluation
File "/home/usr1/Desktop/project/IDA-3D/maskrcnn_benchmark/data/datasets/evaluation/kitti/kitti_eval.py", line 11, in
from maskrcnn_benchmark.structures.boxlist_ops import boxlist_iou
File "/home/usr1/Desktop/project/IDA-3D/maskrcnn_benchmark/structures/boxlist_ops.py", line 6, in
from maskrcnn_benchmark.layers import nms as _box_nms
File "/home/usr1/Desktop/project/IDA-3D/maskrcnn_benchmark/layers/init.py", line 10, in
from .nms import nms
File "/home/usr1/Desktop/project/IDA-3D/maskrcnn_benchmark/layers/nms.py", line 3, in
from maskrcnn_benchmark import _C
ImportError: libcudart.so.11.0: cannot open shared object file: No such file or directory
Hi @swords123
Thank you for making the source code of your paper available. I am trying to reproduce your results, but I am facing a problem.
The server where you have uploaded the pre-processed dataset is really slow. So, I have been unable to download the dataset from the link provided by you. Instead, I went directly to the website of KITTI dataset and downloaded data_object_image_2.zip
and data_object_image_3.zip
. How do I pre-process this data to get it into the format expected by the training script?
hi, thanks for public your code work firslty.
i got a confusion when i compare the code and the paper in the part of orientation regression. according to the paper, the orientation angle was calculated by the estimated viewpoint angle, that makes sense, but when i read the implemented code, it regresses the encoded orientation angle directly instead. so i was wondering did i misunderstand something?
looking forward to your reply, thanks in advance.
Hello, I'm very interested in your work, when can the code be available?
THANK YOU!
Thank you for releasing the code.
I have a question while studying your code, so I want to ask you.
The results of the interference are as follows:
maskrcnn_venchmakr. inference INFO: positions_z : 0.6279727, 0.7480121
0 - 10 : 0.28888804, 3.0077264 0.14014673, 19.665281
10 - 20: 0.xxxxxxx, 0.xxxxxxxxx 0.xxxxxxx, 0.xxxxxxxxx
~
~
~
I wonder what each of the four values after "0-10:" means
Thank you for reading
Hi, I notice that your provide processed dataset and give a data folder format. I am confused about the 'label_3d'? How to get this file from the original KITTI 3D Benchmark? Please provide the data_generate codebase if it is generated by yourself.
Hi,Thank you very much for your excellent work. An error occurred when running "singularity shell --nv ubutu18-cuda10.simg"
~$ singularity shell --nv ubutu18-cuda10.simg
ERROR : Image path doesn't exists
ABORT : Retval = 255
Do you know what might be the reason?
Hello,
I am trying to use the pre-trained model for the evaluation.
./test.sh and python generate_results.py were done without any error.
However, when I run the evaluation with ./kitti_eval/evaluate_object datasets/kitti/label_2 self_exp/exp_1/kitti_test/result_xxx command, it gives me an message like below:
mkdir: cannot create directory 'self_exp/exp_1/kitti_test/result_xxx/plot': No such file or directory
Loading detections...
number of files for evaluation: 0
done.
It fails to make the directory and the number of files for evaluation comes out as 0.
How to solve the issue?
Thank you for your help.
Hello, I use TITAN XP single card to run your code. I followed your instructions strictly. I just delete -m torch.distributed.launch --nproc_per_node=2
in train.sh and test.sh file. But the results I got are lower from those mentioned in your paper. The result are produced from the model you provide. I don't know what went wrong, so I want to ask you.
The result is as follows:
car_detection AP: 90.297791 87.596001 78.844795
car_orientation AP: 90.119148 86.847435 77.664543
car_detection_ground AP: 69.266022 50.099205 43.286926
car_detection_3d AP: 53.662960 36.799820 32.361221
大神有Singularity安装教程吗?分享一下
Excuse me!
When I downloaded processed dataset, I will get the permission deny.
Might you check the link of processed dataset or provide new link?
I'm very interested in your work.
Thank you very much.
Hi, I cant find the part of your code where you construct the cost volume and the nonuniform quantization. Can you point me to where can I find them in your repo?
Thank you very much
我运行train。。。为什么总是出现这个问题ImportError: cannot import name '_C',怎么解决的?
Hi,
I followed your link to the post-processed data, but there are only
where can i find the
in the right format?
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.