ryanjcooper / emnist Goto Github PK

A project designed to explore CNN and the effectiveness of RCNN on classifying the EMNIST dataset.

License: MIT License

Python 68.58% JavaScript 17.81% CSS 3.46% HTML 10.16%

convolutional-neural-networks recurrent-neural-networks emnist mnist flask keras tensorflow learning python3 pythonanywhere

emnist's Introduction

EMNIST

Developed by @coopss

Description

This project was intended to explore the properties of convolution neural networks (CNN) and see how they compare to recurrent convolution neural networks (RCNN). This was inspired by a paper I read that details the effectiveness of RCNNs in object recognition as they perform or even out perform their CNN counterparts with fewer parameters. Aside from exploring CNN/RCNN effectiveness, I built a simple interface to test the more challenging EMNIST dataset dataset (as opposed to the MNIST dataset)

Current Implementation

Multistack CNN
Web-applet testing environment
- Touch screen compatible
- Works best when letter takes up a good portion of the canvas
Read in .mat file
Currently training on the byclass dataset (direct download link)
- See paper for more info

Todo

Update gif with new webapp
Train more models
- RCNN
- Optimize hyperparameters
- Add a noise (gaussian or likewise) layer to input in an attempt to boost accuracy
Move webapp to a host service like PythonAnywhere

Environment

Anaconda: Python 3.5.3

Tensorflow or tensorflow-gpu (See here for more info)
Keras
Flask
Numpy
Scipy

Note: All dependencies for current build can be found in dependencies.txt

Usage

training.py

A training program for classifying the EMNIST dataset

usage: training.py [-h] --file [--width WIDTH] [--height HEIGHT] [--max MAX] [--epochs EPOCHS] [--verbose]

Required Arguments:

-f FILE, --file FILE  Path .mat file data

Optional Arguments

-h, --help            show this help message and exit
--width WIDTH         Width of the images
--height HEIGHT       Height of the images
--max MAX             Max amount of data to use
--epochs EPOCHS       Number of epochs to train on
--verbose         Enables verbose printing

server.py

A webapp for testing models generated from training.py on the EMNIST dataset

usage: server.py [-h] [--bin BIN] [--host HOST] [--port PORT]

Optional Arguments:

-h, --help   show this help message and exit
--bin BIN    Directory to the bin containing the model yaml and model h5 files
--host HOST  The host to run the flask server on
--port PORT  The port to run the flask server on

emnist's People

Contributors

Stargazers

Watchers

Forkers

rdansro djakaitis louisgv learn365 crazzle tahirahmad2030 chuang11 8detect shubhampachori12110095 shahariarrabby turgunyusuf paulgv smscryptor wchzh alexblack2202 diaoxinqiang ygg95 rdeshpande83 shuheng-liu stytchiz zhengyun1008 fitrialif singularitydigitaltechnologies bkbao akeemallen zhangzhdg fabrizziosoares evandrocapo hellkat02 alexhiles py206 cjonestsu manikant92 dxsim ojaswini-nimbalkar tonylvan redsnake314

emnist's Issues

[NON-ISSUE] Cross validating training result

Hello,

I would like to cross validate my loss and accuracy with other. Below is my result:

Can you help me with the usage?

Hi. firstly, thankyou for the wonderful work to test MNIST clissifier model that trained by keras.

but I dont get to how to run "trining.py" to get model.
I cloned your git rep and download "data/mnist_all.mat" from "http://www.cs.nyu.edu/~roweis/data.html"
and give to --file option like above.

python3 training.py --file ./mnist_all.mat

but the log sais

  File "training.py", line 57, in load_data
    mapping = {kv[0]:kv[1:][0] for kv in mat['dataset'][0][0][2]}
KeyError: 'dataset'

looks like the .mat file data is wrong.

can you tell me what .mat file I have to use and explain more about usage?

thank you

No such file or directory: 'bin/mapping.p'

When running training.py, it looks for a file in 'bin/', which as far as I can tell, doesn't exist. My assumption is that on your local repo there was a ./bin/ directory but git ignored it when pushing?

No such file or directory: 'bin/model.yaml'

HI i am getting this issue, please help me, as i'm new to ML so can anyone explains me what is Yaml file and what is the use of this??

Lowercase letters.

Hey,

Is there a way to extract lowercase letters alone? I noticed that they are merged with its uppercase counterparts. I'm working with "letters" dataset.

Thanks in advance.

Pretrained model files

Do you have trained model files you could share? I just wanted to test the model and it would be great if you could share the model files.

IndexError: index 26 is out of bounds for axis 1 with size 26

Dear Ryan Cooper, I faced an issue when I try to train the letters only dataset emnist-letters.mat when I run the training - I get the below error - can you help?