From the course: Transfer Learning for Images Using PyTorch: Essential Training

Unlock the full course today

Join today to access over 22,600 courses taught by industry experts or purchase this course individually.

What is transfer learning?

What is transfer learning?

From the course: Transfer Learning for Images Using PyTorch: Essential Training

Start my 1-month free trial

What is transfer learning?

- [Instructor] Let's say you had an object detection challenge. One might be to tell the difference between, say, different kinds of fruit, and another object detection challenge might be to detect, say, people, potted plants, and cups. Now, you might be tempted to think that for the fruit detection challenge, we just need to gather a whole load of different images of fruit and train our model, and then you do exactly the same thing for the object detection with people, potted plants, and cups, except this around you gather a whole load of images with people, potted plants, and cups and train that model. And that's definitely one way to go about things, but transfer learning offers you a new approach. What you could do instead is to train your model with images for the different types of fruit, and then you use exactly the same model to detect between people, potted plants, and cups, but retrain only the final classifier…

Contents