Introduction to Attention-Based Neural Networks

Course details

Attention-based models allow neural networks to focus on the most important features of the input, thus producing better results at the output. In this course, Janani Ravi explains how recurrent neural networks work and builds and trains two image captioning models one without attention and another using attention models and compares their results. If you have some experience and understanding of how neural networks work and want to see what attention-based models can do for you, check out this course.

Concepts

0. Introduction

01 - Prerequisites
02 - What are attention-based models
03 - Attention in language generation and translation models

1. Recurrent Neural Networks to Learn Sequential Data

04 - Feed forward networks and their limitations
05 - Recurrent neural networks for sequential data
06 - The need for long memory cells
07 - LSTM and GRU cells
08 - Types of RRNNS

2. Encoder-Decoder Networks for Language Models

09 - Language generation models
10 - Sequence to sequence models for language translation

3. Attention-Based Neural Networks

11 - The role of attention in sequence to sequence models
12 - Attention mechanism in sequence to sequence models
13 - Alignment weights in attention models
14 - Bahdanau attention
15 - Attention models for image captioning
16 - Encoder decoder structure for image captioning

4. Image Captioning Model without Attention

17 - Setting up Colab and Google Drive
18 - Loading in the Flickr8k dataset
19 - Constructing the vocabulary
20 - Setting up the dataset class
21 - Implementing utility functions for training data
22 - Building the encoder CNN
23 - Building the decoder RNN
24 - Setting up the sequence to sequence model
25 - Training the image captioning model

5. Image Captioning Model Using Attention

26 - Loading the dataset and setting up utility functions
27 - The encoder CNN generating unrolled feature maps
28 - Implementing Bahdanau attention
29 - The decoder RNN using attention
30 - Generating captions using attention
31 - Training the attention-based image captioning model
32 - Visualizing the model's attention

Conclusion

33 - Summary and next steps