sbodenstein

-- 

require 'cudnn'
require 'cunn'

torch.setdefaulttensortype('torch.FloatTensor')

-- Get weight dim
function checkSums(rnn, seqLength, batch, inputDim, hiddenSize, layerNum, bidirectional)
	rnn:reset()

dannguyen

Using Python 3 + Google Cloud Vision API's OCR to extract text from photos and scanned documents

Just a quickie test in Python 3 (using Requests) to see if Google Cloud Vision can be used to effectively OCR a scanned data table and preserve its structure, in the way that products such as ABBYY FineReader can OCR an image and provide Excel-ready output.

The short answer: No. While Cloud Vision provides bounding polygon coordinates in its output, it doesn't provide it at the word or region level, which would be needed to then calculate the data delimiters.

On the other hand, the OCR quality is pretty good, if you just need to identify text anywhere in an image, without regards to its physical coordinates. I've included two examples:

####### 1. A low-resolution photo of road signs

kbroman

#include <Rcpp.h>
using namespace Rcpp;

// [[Rcpp::export]]
void r_message(std::string text)
{
    Rcpp::Function msg("message");
    msg(text);
}

romainfrancois

#include <Rcpp.h>

// thanks to Simon for the macro magic
// problem #1: R internals uses "private" and "class" as 
//             member names. So a C++ compiler is confused
//             when we include Connections.h
//             this workaround uses the preprocessor to rename
//             class as class_name and private as private_ptr
#define class class_name
#define private private_ptr

BenLangmead

class DeBruijnGraph:
    """ A de Bruijn multigraph built from a collection of strings.
        User supplies strings and k-mer length k.  Nodes of the de
        Bruijn graph are k-1-mers and edges correspond to the k-mer
        that joins a left k-1-mer to a right k-1-mer. """

    @staticmethod
    def chop(st, k):
        """ Chop a string up into k mers of given length """
        for i in xrange(0, len(st)-(k-1)):

pssguy

# load requisite libraries
library(shiny)
library(googleVis) 

# load raw data (NB not provided)
tableByGame <- read.csv("../tableByGame.csv",stringsAsFactors=FALSE)

# to create chart need to repeat one column and get negative of league position as hack
tableByGame$game <- tableByGame$seasonGame
tableByGame$lgPos <- -tableByGame$lgPos

yledu

function writeStuff(n)
    f=open("data.txt","w")
    for el in 1:n
        r=randi(Int32)
        println(f,r)
    end 
    close(f)
end

N = 1000_000

jcheng5

library(shiny)
library(ggplot2)

function(input, output) {

  dataset <- reactive({
    diamonds[sample(nrow(diamonds), input$sampleSize),]
  })

  output$plot <- renderPlot({

smc77

num.iterations <- 1000

# Download South African heart disease data
sa.heart <- read.table("http://www-stat.stanford.edu/~tibs/ElemStatLearn/datasets/SAheart.data", sep=",",head=T,row.names=1)

x <- sa.heart[,c("age", "ldl")]
y <- sa.heart$chd
plot(x, pch=21, bg=c("red","green")[factor(y)])

# Function to standardize input values

tc

package imagej;

import ij.*;
import ij.io.*;
import ij.process.*;
import java.io.*;
import javax.imageio.*;
import javax.imageio.stream.*;
import java.awt.image.BufferedImage;