How to render results in image and BEV

#1

given the results.json file,

“results”: {
“3e8750f331d7499e9b5123e9eb70f2e2”: [
{
“sample_token”: “3e8750f331d7499e9b5123e9eb70f2e2”,
“translation”: [
583.2911376953125,
1656.5841064453125,
1.8910058736801147
],
“size”: [
1.8930909633636475,
4.460857391357422,
1.6006189584732056
],
“rotation”: [
-0.6928867994995735,
0.6779766127186685,
-0.18024127824514968,
0.16663936300380147
],
“velocity”: [
7.791160583496094,
-4.21278715133667
],
“detection_name”: “car”,
“attribute_name”: “vehicle.moving”,
“detection_score”: 0.7195667028427124,
“tracking_name”: “car”,
“tracking_score”: 0.7195667028427124,
“tracking_id”: 1,
“sensor_id”: 4,
“det_id”: -1
} ]
}

how can I render the bounding box per object in a specific scene/sample.

#2

@mfaqihts you simply need to modify this line in nuscenes-devkit to load the inferences from your results.json: https://github.com/nutonomy/nuscenes-devkit/blob/28765b8477dbd3331bacd922fada867c2c4db1d7/python-sdk/nuscenes/nuscenes.py#L315

1 Like