It depends on your model but not the video. Please refer to DeepStream SDK FAQ - #21 by mchi for the accuracy issues.
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| Inference from very wide source | 9 | 636 | October 12, 2021 | |
| what is shape of converter's output with two video | 12 | 1246 | March 12, 2018 | |
| Input Resolution of default deepstream caffe model | 6 | 694 | October 12, 2021 | |
| Accuracy decreased when using deepstream pipeline | 8 | 1600 | August 28, 2021 | |
| TLT detectnet_v2 set training width and height | 16 | 973 | October 12, 2021 | |
| DeepStream GST-NVSTREAMMUX change width and height doesn't affect FPS | 12 | 6857 | October 12, 2021 | |
| Low accuracy nvinfer + nvtracker | 3 | 555 | October 12, 2021 | |
| DeepStream 2.0 Can Run in 1080p or 640p ？ | 7 | 702 | October 12, 2021 | |
| resnet10.caffemodel_b8_fp16.engine is optimized for | 10 | 1444 | October 12, 2021 | |
| Finding inaccurate result while testing model(TLT trained model) with deepstream | 14 | 1155 | October 12, 2021 |