[feature] Initial release of Yolo v3 node

[amc-nu]

Status

PRODUCTION / DEVELOPMENT

Description

This nodes implements yolo v3 in Autoware. autowarefoundation/autoware_ai#116

Todos

• Tests
  • CUDA 8.0
  • CUDA 9.0
• Documentation

Steps to Test or Reproduce

1. Checkout this branch.
2. Follow instruction in README.md in the package.
3. Compile ./catkin_make_release
4. Download the pretrained model
5. Pretrained model will be searched by default inside /darknet/data/ folder in the package. If the file is located in a different path, use the pretrained_model_file param in the launch file. Or select it using RTM
6. Launch the node
7. Play a rosbag with image data
8. Open RVIZ and add Image Viewer
9. Detections are published in /obj_car/image_obj and /obj_person/image_obj

[kitsukawa]

@amc-nu Where can I get the model file for Yolo V3?

[amc-nu]

@kitsukawa
From the original author website:
https://pjreddie.com/media/files/yolov3.weights

[kitsukawa]

@amc-nu When launching the node, the error CUDA Error: unknown error: Permission denied occurs. Do you know the solution for this?

[amc-nu]

what cuda version are you using?
can you run nvidia-smi?

[kitsukawa]

CUDA 8.0

Mon Apr  9 13:49:12 2018       
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 390.25                 Driver Version: 390.25                    |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|===============================+======================+======================|
|   0  GeForce GTX TIT...  Off  | 00000000:01:00.0  On |                  N/A |
| 28%   41C    P8    30W / 250W |    580MiB /  6074MiB |     15%      Default |
+-------------------------------+----------------------+----------------------+
                                                                               
+-----------------------------------------------------------------------------+
| Processes:                                                       GPU Memory |
|  GPU       PID   Type   Process name                             Usage      |
|=============================================================================|
|    0      1093      G   /usr/bin/X                                   389MiB |
|    0      2244      G   compiz                                        99MiB |
|    0      4016      G   ...-token=E8C11D28B4461DB4E08B5771EBF6C094    54MiB |
|    0      4989      G   /opt/ros/indigo/lib/rviz/rviz                 30MiB |

[amc-nu]

can you also try to run deviceQuery app from CUDA examples?

[kitsukawa]

./deviceQuery Starting...

 CUDA Device Query (Runtime API) version (CUDART static linking)

Detected 1 CUDA Capable device(s)

Device 0: "GeForce GTX TITAN Black"
  CUDA Driver Version / Runtime Version          9.1 / 8.0
  CUDA Capability Major/Minor version number:    3.5
  Total amount of global memory:                 6075 MBytes (6369771520 bytes)
  (15) Multiprocessors, (192) CUDA Cores/MP:     2880 CUDA Cores
  GPU Max Clock rate:                            980 MHz (0.98 GHz)
  Memory Clock rate:                             3500 Mhz
  Memory Bus Width:                              384-bit
  L2 Cache Size:                                 1572864 bytes
  Maximum Texture Dimension Size (x,y,z)         1D=(65536), 2D=(65536, 65536), 3D=(4096, 4096, 4096)
  Maximum Layered 1D Texture Size, (num) layers  1D=(16384), 2048 layers
  Maximum Layered 2D Texture Size, (num) layers  2D=(16384, 16384), 2048 layers
  Total amount of constant memory:               65536 bytes
  Total amount of shared memory per block:       49152 bytes
  Total number of registers available per block: 65536
  Warp size:                                     32
  Maximum number of threads per multiprocessor:  2048
  Maximum number of threads per block:           1024
  Max dimension size of a thread block (x,y,z): (1024, 1024, 64)
  Max dimension size of a grid size    (x,y,z): (2147483647, 65535, 65535)
  Maximum memory pitch:                          2147483647 bytes
  Texture alignment:                             512 bytes
  Concurrent copy and kernel execution:          Yes with 1 copy engine(s)
  Run time limit on kernels:                     Yes
  Integrated GPU sharing Host Memory:            No
  Support host page-locked memory mapping:       Yes
  Alignment requirement for Surfaces:            Yes
  Device has ECC support:                        Disabled
  Device supports Unified Addressing (UVA):      Yes
  Device PCI Domain ID / Bus ID / location ID:   0 / 1 / 0
  Compute Mode:
     < Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >

deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 9.1, CUDA Runtime Version = 8.0, NumDevs = 1, Device0 = GeForce GTX TITAN Black
Result = PASS

[amc-nu]

can you please show the error log after launching the node?

[kitsukawa]

['roslaunch', 'yolo3_detector', 'yolo3.launch', 'score_threshold:=0.5', 'nms_threshold:=0.45', 'image_src:=/image_raw', 'network_definition_file:=/home/kitsukawa/Autoware/ros/src/computing/perception/detection/packages/yolo3_detector/darknet/cfg/yolov3.cfg', 'pretrained_model_file:=/home/kitsukawa/Autoware/ros/src/computing/perception/detection/packages/yolo3_detector/darknet/data/yolov3.weights', 'gpu_device_id:=0', 'camera_id:=']
pid=9701
sched policy=OTHER prio=0
... logging to /home/kitsukawa/.ros/log/37bcbb22-3bae-11e8-9cb3-d0509946b845/roslaunch-lab-9701.log
Checking log directory for disk usage. This may take awhile.
Press Ctrl-C to interrupt
Done checking log file disk usage. Usage is <1GB.

started roslaunch server http://lab:49728/

SUMMARY
========

PARAMETERS
* /rosdistro: indigo
* /rosversion: 1.11.21
* /yolo3_detector/gpu_device_id: 0
* /yolo3_detector/image_raw_node: //image_raw
* /yolo3_detector/network_definition_file: /home/kitsukawa/A...
* /yolo3_detector/nms_threshold: 0.45
* /yolo3_detector/pretrained_model_file: /home/kitsukawa/A...
* /yolo3_detector/score_threshold: 0.5

NODES
/
yolo3_detector (yolo3_detector/yolo3_detector)

ROS_MASTER_URI=http://localhost:11311

core service [/rosout] found
layer     filters    size              input                output
0 CUDA Error: unknown error: Permission denied
[ INFO] [1523249751.328316972]: Setting image node to //image_raw
[ INFO] [1523249751.328619999]: Network Definition File (Config): /home/kitsukawa/Autoware/ros/src/computing/perception/detection/packages/yolo3_detector/darknet/cfg/yolov3.cfg
[ INFO] [1523249751.328973700]: Pretrained Model File (Weights): /home/kitsukawa/Autoware/ros/src/computing/perception/detection/packages/yolo3_detector/darknet/data/yolov3.weights
[ INFO] [1523249751.329615928]: [yolo3_detector] score_threshold: 0.500000
[ INFO] [1523249751.330281136]: [yolo3_detector] nms_threshold: 0.450000
[ INFO] [1523249751.330293774]: Initializing Yolo3 on Darknet...
CUDA Error: unknown error
[yolo3_detector-1] process has died [pid 9758, exit code 255, cmd /home/kitsukawa/Autoware/ros/devel/lib/yolo3_detector/yolo3_detector __name:=yolo3_detector __log:=/home/kitsukawa/.ros/log/37bcbb22-3bae-11e8-9cb3-d0509946b845/yolo3_detector-1.log].
log file: /home/kitsukawa/.ros/log/37bcbb22-3bae-11e8-9cb3-d0509946b845/yolo3_detector-1*.log
process[yolo3_detector-1]: started with pid [9758]
all processes on machine have died, roslaunch will exit
shutting down processing monitor...
... shutting down processing monitor complete
done

[Mor-harry]

@kitsukawa @amc-nu launching the node, the error CUDA Error: unknown error: Permission denied occurs. Do you know the solution for this?I still have this problem, in the latest version of 1.7.0.