I’m having the same issue: using the default jupyter notebook and kitti dataset my tlt-infer works great, but the deepstream detection is very poor. See this post Can't get TLT trained model get to work on Deepstream - Jetson (NX) - #2 by kayccc. Tried the decectnet config file, tried all possible eps, dbscan settings. And yes, I converted it on the xavier nx.
Related topics
Topic | Replies | Views | Activity | |
---|---|---|---|---|
TLT trained model accuracy worse after deployment | 11 | 830 | October 12, 2021 | |
Label files generated by tlt-infer | 10 | 827 | October 12, 2021 | |
Little to no detection on Deepstream-App compared to TLT's infer using the same model | 6 | 628 | October 12, 2021 | |
Models trained using TLT perform considerably worse when deployed on DeepStream | 6 | 465 | October 12, 2021 | |
Multiple Bounding Boxes for the Same Detection | 5 | 1008 | May 6, 2022 | |
Issue with image classification tutorial and testing with deepstream-app | 34 | 5780 | October 12, 2021 | |
Deepstream with tlt resnet50 model giving unknown warning | 11 | 688 | October 12, 2021 | |
Negative confidence in DP 6.1.1 | 7 | 448 | September 22, 2023 | |
Converting Custom RetinaNet model to TensorRT in DeepStream | 28 | 82 | January 21, 2025 | |
Integrating Tao Models (detectnet_v2) into Deepstream SDK | 11 | 967 | March 24, 2023 |