Hi, good day. I would like to ask if anyone here is able to associate tracklets or detection results with their recognitions using a script node? We were able to track the detections and decode the recognition data using pure Python, but we were having a hard time associating them. Is this still possible or not? Any tips or advice are much appreciated.
The code below is the MRE, where we try to use TwoStageHostSync to associate it, but it doesn't work. It outputs in a second, then goes empty or none.
Best regards,
MarcSue
queues = {}
age = 0
gender_str = 0
width_value = 0
height_value = 0
# node.warn('i am here before while')
sync = TwoStageHostSeqSync()
while True:
try:
for name in ["tracklets", "age_gender"]:
queues[name] = node.io[name].tryGet()
for name, q in queues.items():
# Add all msgs (color frames, object detections and recognitions) to the Sync class.
if q:
msg = q
# node.warn(f"msg:{msg}")
sync.add_msg(msg, name)
msgs = sync.get_msgs()
if msgs is not None:
tracklets = msgs["tracklets"].tracklets
age_gender = msgs["age_gender"]
width = 300
height = 300
for i, tracklet in enumerate(tracklets):
face_id = tracklet.id
recogs = age_gender[i]
age_fp16 = recogs.getLayerFp16('age_conv3')
if isinstance(age_fp16, list):