This document discusses the building blocks of neural networks. It identifies 5 key building blocks: 1) dense connections, 2) embeddings for processing text, 3) tied weights for reducing parameters, 4) recurrence for handling sequences, and 5) convolutions for processing images. It provides examples of how each building block can be used, such as using embeddings to represent words, recurrence to handle sequential data like text, and convolutions to process image data. The document argues that these building blocks can be combined in different ways to solve a variety of problems, such as predicting the next word in a sequence, detecting objects in images, or generating image captions.
51. BUILDING BLOCKS
1. dense connections
2. embeddings for processing text
3. tied weights for reducing parameters
4. recurrence for handling sequences
60. BUILDING BLOCKS
1. dense connections
2. embeddings for processing text
3. tied weights for reducing parameters
4. recurrence for handling sequences
5. convolutions for processing images
62. PROBLEM: CAPTIONING IMAGES
Show and Tell: A Neural Image Caption Generator: Vinyals et al
A group of young
people playing a
game of frisbee.
output:input: