This repo contains a number of standard deep learning models that can be run with the neon libraries. The specific model type subdirectories include more detailed information on the various models. Below is a summary of the various tasks and models included here that address those tasks.

Image classification

ImageNet

• Googlenet
• VGG
• Alexnet

CIFAR-10

• Deep Residual Network
• All CNN

Object localization

• Fast-RCNN

Scene Classification

• Deep Residual Network

Image Captioning

• LSTM

Video Games

• Deep Q-network

NLP

bAbI (Question Answering)

• GRU/LSTM model

Sentiment classification

• LSTM (IMDB dataset)

Video

• C3D (UCF101 dataset)

Transfer Learning

• Classification using multi scale sampling (PASCALVOC dataset)

Speech

• Deep Speech 2 (Librispeech dataset)