/
andrea-benericetti-machine-learning-approaches-for-road-scene-video-analysis.json
34 lines (34 loc) · 1.75 KB
/
andrea-benericetti-machine-learning-approaches-for-road-scene-video-analysis.json
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
{
"copyright_text": null,
"description": "Dashboard cameras (dashcams), inward or front-facing cameras installed\nin personal or commercial vehicles, are becoming increasingly popular\ndue to the pervasiveness of their applications: driver safety,\nautonomous driving, fleet management systems, insurance, theft detection\nare some examples. Focusing on safety, one of the main problems is to\nanalyze videos and automatically detect dangerous situations occurring\nin them, such as the risk of a near crash with another vehicle or\npedestrian. In this talk, we show how we tackle this real- world problem\nat Verizon Connect, using a mix of state-of-the-art deep learning\nmethods and traditional computer vision / machine learning techniques,\nleveraging libraries such as scikit-learn, scipy/numpy, pandas, openCV,\nand keras/tensorflow. We also describe how we use AWS and docker to\ndeploy, serve and scale the application to customers all over the world.\n\n**Feedback form:** https://python.it/feedback-1603\n\nin \\_\\_on **Friday 3 May** at 11:15 `**See\nschedule** </en/sprints/schedule/pycon10/>`__\n",
"duration": 1953,
"language": "eng",
"recorded": "2019-05-03",
"related_urls": [
{
"label": "Conference schedule",
"url": "https://www.pycon.it/p3/schedule/pycon10/"
}
],
"speakers": [
"Andrea Benericetti"
],
"tags": [
"ComputerVision",
"analytics",
"scikit-learn",
"opencv",
"video",
"machine-learning",
"optical-flow",
"pandas"
],
"thumbnail_url": "https://i.ytimg.com/vi/CmzI17zWcmA/maxresdefault.jpg",
"title": "Machine learning approaches for road scene video analysis",
"videos": [
{
"type": "youtube",
"url": "https://www.youtube.com/watch?v=CmzI17zWcmA"
}
]
}