Bounding Boxes Shifted After TRT Inference vs Manually Drawn KITTI Labels — Cause?

Accelerated Computing Intelligent Video Analytics TAO Toolkit
1

Please provide the following information when requesting support.

• Hardware (T4/V100/Xavier/Nano/etc) Nvidia A600

• Network Type (Detectnet_v2/Faster_rcnn/Yolo_v4/LPRnet/Mask_rcnn/Classification/etc) Yolo_v4_tiny

• TLT Version (Please run “tlt info --verbose” and share “docker_tag” here) 5.3.0
• Training spec file(If have, please share here)

import os
%env USER_EXPERIMENT_DIR=/workspace/tao-experiments/yolo_v4_tiny
%env DATA_DOWNLOAD_DIR=/workspace/tao-experiments/data
%env LOCAL_PROJECT_DIR=/home/orkohunter/pk_iteration_04
os.environ[“LOCAL_DATA_DIR”] = os.path.join(os.getenv(“LOCAL_PROJECT_DIR”, os.getcwd()), “data”)
os.environ[“LOCAL_EXPERIMENT_DIR”] = os.path.join(os.getenv(“LOCAL_PROJECT_DIR”, os.getcwd()), “yolo_v4_tiny”)

The sample spec files are present in the same path as the downloaded samples.

os.environ[“LOCAL_SPECS_DIR”] = os.path.join(
os.getenv(“NOTEBOOK_ROOT”, os.getcwd()),
“specs”
)
%env SPECS_DIR=/workspace/tao-experiments/yolo_v4_tiny/specs

Showing list of specification files.

!ls -rlt $LOCAL_SPECS_DIR

• How to reproduce the issue ? (This is for errors. Please share the command line and the detailed log here.)

2

May I know more details about the issue you meet? More, is it fine when you run !tao model yolo_v4_tiny inference in tao_tutorials/notebooks/tao_launcher_starter_kit/yolo_v4_tiny/yolo_v4_tiny.ipynb at main · NVIDIA/tao_tutorials · GitHub?