stecman

Controlling the Brother P-Touch Cube label maker from a computer

The Brother PTP300BT label maker is intended to be controlled using the official Brother P-Touch Design & Print iOS/Android app. The app has arbitrary limits on what you can print (1 text object and up to 3 preset icons), so I thought it would be a fun challenge to reverse engineer the protocol to print whatever I wanted.

Python code at the bottom if you want to skip the fine details.

Process

Intitially I had a quick peek at the Android APK to see if there was any useful information inside. The code that handles the communication with the printer in Print&Design turned out to be a native library, but the app clearly prepares a bitmap image and passes it to this native library for printing. Bitmaps are definitely something we can work with.

rtqichen

## Weight norm is now added to pytorch as a pre-hook, so use that instead :)

import torch
import torch.nn as nn
from torch.nn import Parameter
from functools import wraps

class WeightNorm(nn.Module):
    append_g = '_g'
    append_v = '_v'

kylemcdonald

import numpy as np
from noise import pnoise2, pnoise3
from scipy.ndimage.filters import gaussian_filter, gaussian_filter1d

def center_and_scale(drawing):
    """
    Translate an entire drawing to the mean location of the points,
    then scale the drawing to fit within +/-1.
    """
    all_points = np.vstack(drawing)

volkancirik

"""
TreeLSTM[1] implementation in Pytorch

Based on dynet benchmarks : 
  https://github.com/neulab/dynet-benchmark/blob/master/dynet-py/treenn.py
  https://github.com/neulab/dynet-benchmark/blob/master/chainer/treenn.py
Other References:
  https://github.com/pytorch/examples/tree/master/word_language_model
  https://github.com/pfnet/chainer/blob/29c67fe1f2140fa8637201505b4c5e8556fad809/chainer/functions/activation/slstm.py
  https://github.com/stanfordnlp/treelstm

ndronen

#!/usr/bin/env python
"""
A quick, partial implementation of ENet (https://arxiv.org/abs/1606.02147) using PyTorch.
The original Torch ENet implementation can process a 480x360 image in ~12 ms (on a P2 AWS
instance).  TensorFlow takes ~35 ms.  The PyTorch implementation takes ~25 ms, an improvement
over TensorFlow, but worse than the original Torch.
"""

from __future__ import absolute_import

farrajota

require 'nn'
 
-- you just need to provide the linear module you want to convert,
-- and the dimensions of the field of view of the linear layer

function convertLinear2Conv1x1(linmodule,in_size)
--[[
    Convert Linear modules to convolution modules. 
    
    Arguments

baraldilorenzo

##VGG16 model for Keras

This is the Keras model of the 16-layer network used by the VGG team in the ILSVRC-2014 competition.

It has been obtained by directly converting the Caffe model provived by the authors.

Details about the network architecture can be found in the following arXiv paper:

Very Deep Convolutional Networks for Large-Scale Image Recognition

K. Simonyan, A. Zisserman

karpathy

"""
Minimal character-level Vanilla RNN model. Written by Andrej Karpathy (@karpathy)
BSD License
"""
import numpy as np

# data I/O
data = open('input.txt', 'r').read() # should be simple plain text file
chars = list(set(data))
data_size, vocab_size = len(data), len(chars)

lazywei

#!/bin/sh


# this script does absolutely ZERO error checking.   however, it worked
# for me on a RHEL 6.3 machine on 2012-08-08.  clearly, the version numbers
# and/or URLs should be made variables.  cheers,  zmil...@cs.wisc.edu


mkdir mosh
cd mosh

lelandbatey

Description

This simple script will take a picture of a whiteboard and use parts of the ImageMagick library with sane defaults to clean it up tremendously.

The script is here:

#!/bin/bash
convert "$1" -morphology Convolve DoG:15,100,0 -negate -normalize -blur 0x1 -channel RBG -level 60%,91%,0.1 "$2"

Results