Deep Learning, Transfer Learning and Turtles
- Basic familiarity with machine learning
- Basic familiarity with python
In this course, we will use transfer learning and Meeshkan to teach a machine to tell the difference between Teenage Mutant Ninja Turtles and Koopa Troopas with over 95% accuracy. We will then apply this turtle-detector to real turtle images in order to classify if the real turtles look more like TMNTs or Koopa Troopas. You'll be amazed by the results!
Who this course is for:
- Beginner python developers curious about transfer learning
- Users of Slack
- 00:06Online materials for this course
- 02:40Getting your Meeshkan API Key
- 02:36Joining the Meeshkan Community on Slack
- 03:00Setting up your virtual environment
- 04:14Gathering your image data
- 03:47Preprocessing your image data
- 10:27Training using transfer learning
- 02:17Meeshkan CLI
- 04:23Predictions with REAL turtles!
Mike Solomon is the C.E.O. of Meeshkan Machine Learning. When he's not learning with or from computers, he spends his time in France directing the Ensemble 101.
Mike is a graduate of Stanford University, Queen's Belfast, the University of Florida and UPMC in Paris.
Before starting with Meeshkan, he worked in the IRCAM, Jongla and Grame doing Computer Music, Audio and AI research and development.