On Step 3: Edit the Inference Configuration File
[property]
gpu-id=0
net-scale-factor=1
tlt-model-key=tlt_encode
tlt-encoded-model=/dli/task/ngc_assets/dashcamnet_vpruned_v1.0
labelfile-path=/dli/task/dashcam_labels.txt
int8-calib-file=/dli/task/empty_calib_file.txt
input-dims=3;720;1280;0
uff-input-blob-name=input_1
batch-size=1
process-mode=1
model-color-format=0
network-mode=0
num-detected-classes=4
interval=0
gie-unique-id=1
output-blob-names=output_bbox/BiasAdd;output_cov/Sigmoid
cluster-mode=0
On step 4. Build and Run DeepStream Pipeline
4.2
#Import necessary libraries
import sys
import gi
gi.require_version(‘Gst’, ‘1.0’)
from gi.repository import GObject, Gst, GLib
from common.bus_call import bus_call
import pyds
def run(input_file_path):
global inference_output
inference_output=
Gst.init(None)
# Create element that will form a pipeline
print("Creating Pipeline")
pipeline=Gst.Pipeline()
source = Gst.ElementFactory.make("filesrc", "file-source")
source.set_property('location', input_file_path)
h264parser = Gst.ElementFactory.make("h264parse", "h264-parser")
decoder = Gst.ElementFactory.make("nvv4l2decoder", "nvv4l2-decoder")
streammux = Gst.ElementFactory.make("nvstreammux", "Stream-muxer")
streammux.set_property('width', 1280) # Set appropriate width
streammux.set_property('height', 720) # Set appropriate height
streammux.set_property('batch-size', 1) # Set appropriate batch size
pgie = Gst.ElementFactory.make("nvinfer", "primary-inference")
pgie.set_property('config-file-path', os.environ['SPEC_FILE'])
nvvidconv1 = Gst.ElementFactory.make("nvvideoconvert", "convertor")
nvosd = Gst.ElementFactory.make("nvdsosd", "onscreendisplay")
nvvidconv2 = Gst.ElementFactory.make("nvvideoconvert", "convertor2")
capsfilter = Gst.ElementFactory.make("capsfilter", "capsfilter")
caps = Gst.Caps.from_string("video/x-raw, format=I420")
capsfilter.set_property("caps", caps)
encoder=Gst.ElementFactory.make("avenc_mpeg4", "encoder")
encoder.set_property("bitrate", 2000000)
sink=Gst.ElementFactory.make("filesink", 'filesink')
sink.set_property('location', 'output.mpeg4')
sink.set_property("sync", 1)
# Add the elements to the pipeline
print("Adding elements to Pipeline")
pipeline.add(source)
pipeline.add(h264parser)
pipeline.add(decoder)
pipeline.add(streammux)
pipeline.add(pgie)
pipeline.add(nvvidconv1)
pipeline.add(nvosd)
pipeline.add(nvvidconv2)
pipeline.add(capsfilter)
pipeline.add(encoder)
pipeline.add(sink)
# Link the elements together
print("Linking elements in the Pipeline")
source.link(h264parser)
h264parser.link(decoder)
decoder.get_static_pad('src').link(streammux.get_request_pad("sink_0"))
streammux.link(pgie) # Link streammux to primary inference
pgie.link(nvvidconv1) # Link primary inference to nvvidconv1
nvvidconv1.link(nvosd) # Link convertor to onscreendisplay
nvosd.link(nvvidconv2)
nvvidconv2.link(capsfilter)
capsfilter.link(encoder)
encoder.link(sink)
# Attach probe to OSD sink pad
osdsinkpad=nvosd.get_static_pad("sink")
osdsinkpad.add_probe(Gst.PadProbeType.BUFFER, osd_sink_pad_buffer_probe, 0)
# Create an event loop and feed gstreamer bus mesages to it
loop=GLib.MainLoop()
bus=pipeline.get_bus()
bus.add_signal_watch()
bus.connect("message", bus_call, loop)
# Start play back and listen to events
print("Starting pipeline")
pipeline.set_state(Gst.State.PLAYING)
try:
loop.run()
except:
pass
pipeline.set_state(Gst.State.NULL)
return inference_output
4.3
def osd_sink_pad_buffer_probe(pad, info, u_data):
gst_buffer=info.get_buffer()
# Retrieve batch metadata from the gst_buffer
batch_meta=pyds.gst_buffer_get_nvds_batch_meta(hash(gst_buffer))
l_frame=batch_meta.frame_meta_list
while l_frame is not None:
# Initially set the tailgate indicator to False for each frame
tailgate=False
try:
frame_meta=pyds.NvDsFrameMeta.cast(l_frame.data)
except StopIteration:
break
frame_number=frame_meta.frame_num
l_obj=frame_meta.obj_meta_list
# Iterate through each object to check its dimension
while l_obj is not None:
try:
obj_meta=pyds.NvDsObjectMeta.cast(l_obj.data)
# If the object meet the criteria then set tailgate indicator to True
obj_bottom=obj_meta.rect_params.top+obj_meta.rect_params.height
if (obj_meta.rect_params.left > FRAME_WIDTH * 0.3) & (obj_bottom > FRAME_HEIGHT * 0.9):
tailgate = True
except StopIteration:
break
try:
l_obj=l_obj.next
except StopIteration:
break
print(f'Analyzing frame {frame_number}', end='\r')
inference_output.append(str(int(tailgate)))
try:
l_frame=l_frame.next
except StopIteration:
break
return Gst.PadProbeReturn.OK
4.4
tailgate_log=run(input_file_path=‘/dli/task/data/assessment_stream.h264’)
with open(‘/dli/task/my_assessment/answer_4.txt’, ‘w’) as f:
f.write(‘\n’.join(tailgate_log))
The output of the last cell (4.4) is :
reating Pipeline
Adding elements to Pipeline
Linking elements in the Pipeline
Starting pipeline
The Problem:
So the issue here is that after a second from “Starting the pipline”, the below popup appears. I tried restarting, stopping the entire assessment and re-execute the notebook, running the lab the next day, etc. Yet nothing worked.
Any ideas… ?
