Access Metadata of YOLOv10 model in deepstream

Accelerated Computing Intelligent Video Analytics DeepStream SDK
, , , ,
1

• Hardware Platform (Jetson / GPU) - Jetson Nano 4GB
• DeepStream Version - 6.0.1
• JetPack Version (valid for Jetson only) - 4.6.1
• TensorRT Version - 8.2
• CUDA runtime version - 10.2
• CUDA driver version - 8.2
• Issue Type( questions, new requirements, bugs) - question
• How to reproduce the issue ?
Config file:
[application]
enable-perf-measurement=1
perf-measurement-interval-sec=5

[tiled-display]
enable=1
rows=1
columns=1
width=1280
height=720
gpu-id=0
nvbuf-memory-type=0

[source0]
enable=1
type=1
camera-id=0
camera-height=480
camera-width=640
camera-fps-n=25
camera-fps-d=1
camera-v4l2-dev-node=0
#drop-frame-interval=5
#cudadec-memtype=1
num-sources=1

[streammux]
gpu-id=0
live-source=1
batch-size=1
nvbuf-memory-type=0
enable-padding=0
width=640
height=480

[primary-gie]
enable=1
gie-unique-id=1
gpu-id=0
config-file=config_infer_primary_yolo11.txt
batch-size=1
labelfile-path=labels.txt
plugin-type=0

[osd]
enable=1
gpu-id=0
border-width=2
text-size=14
text-color=1;1;1;1;
text-bg-color=0.3;0.3;0.3;1
font=Purisa
nvbuf-memory-type=0
display-text=1
display-bbox=1
nvbuf-memory-type=0
show-clock=0
clock-x-offset=800
clock-y-offset=820
clock-text-size=12
clock-color=1;0;0;0

[sink0]
enable=0
type=2
sync=0
gpu-id=0
nvbuf-memory-type=0

[sink1]
enable=1
type=4
sync=0
source-id=0
gpu-id=0
nvbuf-memory-type=0
width=1280
height=720
codec=1
rtsp-port=4444
------------------------------------------------------------------------------------------------------------------
Model file:
[property]
gpu-id=0
net-scale-factor=0.0039215697906911373
model-color-format=0
onnx-file=v11_models/yolo11n.pt.onnx
model-engine-file=model_b1_gpu0_fp32.engine
#int8-calib-file=calib.table
labelfile-path=labels.txt
batch-size=1
network-mode=0
num-detected-classes=80
interval=0
gie-unique-id=1
process-mode=1
network-type=0
cluster-mode=2
maintain-aspect-ratio=1
symmetric-padding=1
#workspace-size=2000
parse-bbox-func-name=NvDsInferParseYolo
#parse-bbox-func-name=NvDsInferParseYoloCuda
custom-lib-path=nvdsinfer_custom_impl_Yolo/libnvdsinfer_custom_impl_Yolo.so
engine-create-func-name=NvDsInferYoloCudaEngineGet

[class-attrs-all]
nms-iou-threshold=0.45
pre-cluster-threshold=0.25
topk=300
------------------------------------------------------------------------------------------------------------------

I have been using a repo for YOLO in deepstream Yolo-Deepstream. I have a config file to access the webcam video and then process with YOLOv10-n model and then display the plotted video and then stream the plotted video using RTSP. Everything is working fine, I don’t know how to access the metadata of the detection model like bbox, class and confidence values. The reason i want them is that i want to give real-time alert like alarm for detection of a particular vehicle for a continous of 40 frames.
Note: My intention is to know how to get the metadata of every corresponding frames. If its not possible in jetson nano its fine. I can go with jetson orin nano or other dGPU.

3

You need to modify our source code. You can refer to the source code below to learn how to get the informations you wanted.

deepstream\sources\apps\sample_apps\deepstream-app\deepstream_app.c
static GstPadProbeReturn
analytics_done_buf_prob (GstPad * pad, GstPadProbeInfo * info, gpointer u_data)
{