A Simple and Flexible Pytorch Video Pipeline

By paulbridger - 2020-09-23

Description

Taking machine learning models into production for video analytics doesn’t have to be hard. A pipeline with reasonable efficiency can be created very quickly just by plugging together the right librar ...

Summary

September 23, 2020 Taking machine learning models into production for video analytics doesn’t have to be hard.
Worthwhile throughput improvements are possible with a little effort.
so we can understand what’s happening end to end.
$ make frames_into_python.pipeline.png As you can see from the Gst.parse_launch call the constructed pipeline is even simpler than the CLI version since we don’t bother with the progressreport element.

Topics

Backend (0.32)
Machine_Learning (0.12)
NLP (0.04)

Similar Articles

MartinoMensio/spacy-dbpedia-spotlight

By GitHub - 2021-02-14

A spaCy wrapper for DBpedia Spotlight. Contribute to MartinoMensio/spacy-dbpedia-spotlight development by creating an account on GitHub.

Accelerating MLOps at Bayer Crop Science with Kubeflow Pipelines and Amazon SageMaker

By Amazon Web Services - 2021-01-05

This is a guest post by the data science team at Bayer Crop Science. Farmers have always collected and evaluated a large amount of data with each growing season: seeds planted, crop protection inputs ...

My Journey to a serverless transformers pipeline on Google Cloud

By huggingface - 2021-03-18

We’re on a journey to advance and democratize artificial intelligence through open source and open science.

Automatically Build and Deploy Your Python Application in 4 Easy Steps

By Medium - 2020-12-30

When you finish a new version of your Python application, how do you build and deploy it? Do you change the version number, upload the new files to production, and be done with it? There are better…

Easy Machine Translation with Machine Learning and HuggingFace Transformers

By MachineCurve - 2021-02-15

Machine Translation with Python and Transformers - learn how to build an easy pipeline for translation and to extend it to more languages.

How to build a fraud detection solution

By Google Cloud Blog - 2021-03-03

In collaboration with our partner Quantiphi, we developed a smart analytics design pattern that enables you to build a scalable real-time fraud detection solution in one hour using serverless, no-ops ...