Deepstream resnet18

When I use the ResNet18.onnx model (which can be divided into 10 categories in total) for level 2 inference on DeepStream, I can only detect one category (black) and cannot detect other categories .Level 1 reasoning is YOLOV8.Here is my configuration file
[property]
gpu-id=0

labelfile-path=/home/nvidia/sby/resnet18_model_onnx/classes.txt
onnx-file=/home/nvidia/sby/resnet18_model_onnx/best_model_1016.onnx
#model-engine-file=/home/nvidia/sby/resnet18_model_onnx/best_model_1016.onnx_b16_gpu0_fp16.engine
net-scale-factor=0.01742919
offsets=114.75;114.75;114.75
network-type=1
#model-file=/opt/nvidia/deepstream/deepstream-6.0/samples/models/Secondary_CarColor/resnet18.caffemodel
#proto-file=/opt/nvidia/deepstream/deepstream-6.0/samples/models/Secondary_CarColor/resnet18.prototxt
#int8-calib-file=/opt/nvidia/deepstream/deepstream-6.0/samples/models/Secondary_CarColor/cal_trt.bin
#model-engine-file=/root/videoBox/common/color_classification/resnet18.caffemodel_b16_gpu0_fp16.engine
#mean-file=/root/videoBox/common/color_classification/mean.ppm
#labelfile-path=/root/videoBox/common/color_classification/labels.txt
#int8-calib-file=/root/videoBox/common/color_classification/cal_trt.bin
#force-implicit-batch-dim=1
batch-size=16

0=FP32 and 1=INT8 mode

network-mode=1
input-object-min-width=8
input-object-min-height=8
process-mode=2
model-color-format=1
gpu-id=0
gie-unique-id=4

operate-on-gie-id=2

operate-on-class-ids=2;3;5;7;
is-classifier=1
num-detected-classes=10
#uff-input-blob-name=input_1
output-blob-names=predictions/Softmax
classifier-async-mode=1
classifier-threshold=0.51

classifier-threshold=0.51

process-mode=2
#scaling-filter=0
#scaling-compute-hw=0
#[class-attrs-all]
pre-cluster-threshold=0.2
topk=20
nms-iou-threshold=0.5

What class does your classification model target?

What does your label file look like? Can you share the model?

Thank you very much. My classification is various colors, and I found that my label file is different from the one provided by the official. After changing it, it will display normally.

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.