Deepstream 5.1 error: Could not find output coverage layer for parsing objects

My first infer engine is peoplenet and second infer engine is faciallandmark. I have deploy two models in deepstream. But it occur this error: “Could not find output coverage layer for parsing objects”. I think some errors in two config files.

pgie.config file:


[property]
gpu-id=0
net-scale-factor=0.0039215697906911373
tlt-model-key=tlt_encode
tlt-encoded-model=/opt/nvidia/deepstream/deepstream-5.1/samples/models/tlt_pretrained_models/peoplenet/resnet34_peoplenet_pruned.etlt
labelfile-path=/opt/nvidia/deepstream/deepstream-5.1/samples/models/tlt_pretrained_models/peoplenet/labels.txt
infer-dims=3;544;960
uff-input-blob-name=input_1
force-implicit-batch-dim=1
batch-size=4
network-type=0
process-mode=1
model-color-format=0
network-mode=1
num-detected-classes=3
interval=0
gie-unique-id=1
output-blob-names=output_bbox/BiasAdd;output_cov/Sigmoid
## 0=Group Rectangles, 1=DBSCAN, 2=NMS, 3= DBSCAN+NMS Hybrid, 4 = None(No clustering)
cluster-mode=3
workspace-size=4096
# filter-out-class-ids=0;1

[class-attrs-all]
pre-cluster-threshold=0.2
eps=0.7
minBoxes=1

#Use the config params below for dbscan clustering mode
[class-attrs-all]
detected-min-w=4
detected-min-h=4
minBoxes=3

## Per class configurations
[class-attrs-0]
pre-cluster-threshold=0.05
eps=0.7
dbscan-min-score=0.95

[class-attrs-1]
pre-cluster-threshold=0.05
eps=0.7
dbscan-min-score=0.5

[class-attrs-2]
pre-cluster-threshold=0.1
eps=0.6
dbscan-min-score=0.95

sgie file:


[property]
gpu-id=0
net-scale-factor=0.0039215697906911373
gie-unique-id=2
operate-on-gie-id=1
operate-on-class-ids=2
is-classifier=1
process-mode=2
force-implicit-batch-dim=1
batch-size=1
# 0=FP32 and 1=INT8 mode
network-mode=1

model-engine-file=./models/FacialLandmarksEstimation/saved.engine
output-blob-names=softargmax/strided_slice:0;softargmax/strided_slice_1:0
model-color-format=2
network-type=0

input-object-min-width=64
input-object-min-height=64
num-detected-classes=1

Can you provide detail reproduce step? So I can try it in my side.

There are three download models include FaceDetect, FacialLandMark and EmotionNet from NGC.
deepstream_mytest.py (17.6 KB)
my_sgie1_config.txt (3.1 KB)
my_sgie2_config.txt (3.2 KB)
my_tracker_config.txt (1.8 KB)
myconfig.txt (3.7 KB)
tracker_config.yml (6.6 KB)

Thanks. I will feedback to you if I get any progress.

I have view sources/libs/nvdsinfer/nvdsinfer_context_impl_out_parsing.cpp. The issue occur in line 59. As shown below


I re-check my first infer config file. I have config this output layer parameters.

whether do I miss some other parameters?

Can you try FacialLandMark samples in below link?

deepstream_tao_apps/apps/tlt_others at release/tlt3.0 · NVIDIA-AI-IOT/deepstream_tao_apps (github.com)