valavanp,
What do you mean it isn't efficient? Perhaps you should look up the definition of the word. I will go out on a limb and guess that you want it to run faster. In any case, converting the CSV was not necessary.

$!/usr/bin/perl
use strict;
use warnings;
use Text::CSV_XS;

my $file = $ARGV[0] or die "Usage: $0 <input_file>";
open(my $fh, '<', $file) or die "Unable to open '$file' for reading: $
+!";

my $csv = Text::CSV_XS->new;
while (<$fh>) {
    chomp;
    if ($csv->parse($_)) {
        printf "'%s'\t'=>\t'%s'\n", $csv->fields;
    }
    else {
        print STDERR "parse() failed for '$_': ", $csv->error_input;
        exit;
    }
}
[download]

Shamelessly stole Corion's printf() since it was better.