Implement Image Captioning with Recurrent Neural Networks
This course will teach you how to build and train image captioning models using TensorFlow, with the help of a case study - building a model for image tagging. You will learn how to prepare the data for model training and evaluate the trained model.
What you'll learn
Manually interpreting billions of images is time-consuming and almost impossible. But if we teach machines to understand images, this task will become much more efficient. In this course, Implement Image Captioning with Recurrent Neural Networks, you’ll learn to build and train image captioning models with RNNs using TensorFlow. First, you’ll explore the broader understanding of recurrent neural networks along with an overview of image captioning and how CNNs can help us to understand images. Next, you’ll discover how to prepare image and text data. Then, you'll learn how to develop a deep learning model for image captioning, and different options to evaluate that model using TensorFlow. Finally, you’ll understand the implementation of the data science process. When you’re finished with this course, you’ll have the skills and knowledge of RNNs and CNNs needed to build image captioning models.
Table of contents
- Overview 1m
- Demo: Build the Attention Model for Image-captioning Using TensorFlow 5m
- Demo: Implement CNN Encoder in TensorFlow 1m
- Demo: Implement RNN Decoder with Attention & Sentence Generator 2m
- Demo: Define the Loss Function and Model Checkpoints 3m
- Demo: Perform Model Training 3m
- Demo: Making Predictions out of the Trained Model 2m
- Summary 1m