Google Cloud Platform1.17 млн
Опубликовано 8 апреля 2021, 16:00
Blog post → goo.gle/3sYS9VV
Cameratrap for backyard use or research → goo.gle/39RvT8z
Notebook to try our codelab → goo.gle/3rXWur4
Camera trap dataset used for training → goo.gle/3wyOMXH
Machine learning plays a crucial role in surveilling and measuring the biodiversity of wildlife in a non-invasive way. Welcome to the first episode of People and Planet AI - a series dedicated to exploring how machine learning helps us learn about the biological variety and variability on earth. In this first episode, we show you how to create an image classification app for wildlife camera traps with the help of Dataflow and Google Cloud’s AI Platform Unified. Watch to learn how you can create an application that tracks wildlife biodiversity!
Chapters:
0:00 - 00:42 - Background of this episode
00:42 - 01:04 - Series ethos
01:04 - 02:52 - Wildlife Insights case study
02:52 - 03:55 - Hardware components
03:55 - 06:20 - Software components to build ML Classification model
06:20 - 06:49 - Examples of how the model performs
06:49 - 06:52 - Uploading model into microcontrollers
06:53 - 07:06 - Uploading new photos in the future
07:06 - 07:18 - Create a species dashboard
07:18 - 07:29 - Recap
07:29 - 07:47 - What to do next
08:01 - 08:21 - Tanya’s fun fact on using camera traps
Leverage Wildlife Insights’s platform:
WildlifeInsights.org → goo.gle/39SIqst
Blog posts on Wildlife Insights:
Project from Google’s Sustainability site → goo.gle/31YdGBS
Machine Learning in Wildlife Insights → goo.gle/3mpbwVy
ML enhancements in Wildlife Insights → goo.gle/3cZpuu6
Google resources:
How to start a Google cloud project → goo.gle/39SJjkN
AI Platform Unified → goo.gle/3fPYmjd
AutoML vision → goo.gle/3rZniH6
Dataflow → goo.gle/3sUIQWO
BigQuery → goo.gle/31WxeGQ
Cloud Storage → goo.gle/31ViIiM
Cloud IoT core → goo.gle/3wA2c5O
Watch more episodes of People and Planet AI → goo.gle/PeopleAndPlanetAI
Subscribe to Google Cloud Tech → goo.gle/GoogleCloudTech
#PeopleAndPlanetAI #MachineLearning
product: AI Platform Unified, Dataflow; fullname: AGV;
Cameratrap for backyard use or research → goo.gle/39RvT8z
Notebook to try our codelab → goo.gle/3rXWur4
Camera trap dataset used for training → goo.gle/3wyOMXH
Machine learning plays a crucial role in surveilling and measuring the biodiversity of wildlife in a non-invasive way. Welcome to the first episode of People and Planet AI - a series dedicated to exploring how machine learning helps us learn about the biological variety and variability on earth. In this first episode, we show you how to create an image classification app for wildlife camera traps with the help of Dataflow and Google Cloud’s AI Platform Unified. Watch to learn how you can create an application that tracks wildlife biodiversity!
Chapters:
0:00 - 00:42 - Background of this episode
00:42 - 01:04 - Series ethos
01:04 - 02:52 - Wildlife Insights case study
02:52 - 03:55 - Hardware components
03:55 - 06:20 - Software components to build ML Classification model
06:20 - 06:49 - Examples of how the model performs
06:49 - 06:52 - Uploading model into microcontrollers
06:53 - 07:06 - Uploading new photos in the future
07:06 - 07:18 - Create a species dashboard
07:18 - 07:29 - Recap
07:29 - 07:47 - What to do next
08:01 - 08:21 - Tanya’s fun fact on using camera traps
Leverage Wildlife Insights’s platform:
WildlifeInsights.org → goo.gle/39SIqst
Blog posts on Wildlife Insights:
Project from Google’s Sustainability site → goo.gle/31YdGBS
Machine Learning in Wildlife Insights → goo.gle/3mpbwVy
ML enhancements in Wildlife Insights → goo.gle/3cZpuu6
Google resources:
How to start a Google cloud project → goo.gle/39SJjkN
AI Platform Unified → goo.gle/3fPYmjd
AutoML vision → goo.gle/3rZniH6
Dataflow → goo.gle/3sUIQWO
BigQuery → goo.gle/31WxeGQ
Cloud Storage → goo.gle/31ViIiM
Cloud IoT core → goo.gle/3wA2c5O
Watch more episodes of People and Planet AI → goo.gle/PeopleAndPlanetAI
Subscribe to Google Cloud Tech → goo.gle/GoogleCloudTech
#PeopleAndPlanetAI #MachineLearning
product: AI Platform Unified, Dataflow; fullname: AGV;
Свежие видео