Linux Tensorflow Script Coursera adaption: handwriting recognition for 1x specific writer (offline usage) - 30/04/2018 12:28 EDT
$10-90 USD
Closed
Posted almost 6 years ago
$10-90 USD
Paid on delivery
=== Background ===
The bidder should create an linux based Tensorflow / Octave or other statistic, neuronal net (?) based handwriting recognition script for one specific writer. It should be linux commandline only. Bash should call Python libraries. The main logic should be in bash. The script should have the possibility to be trained with new training data, so that the script gets better over time.
=== Deliverables ==
Deliverable 1: The script covering needed INPUT / OUTPUT.
Deliverable 2: Training feature with additional training data (HowTo needed how to train)
=== Input ===
a PDF/JPG of handwritten text
=== Output ===
The script (tensorflow, ocatve or other ML tool) should take an PDF as an input and export the recognized text
=== Sources found ===
[1] [login to view URL]
[2] [login to view URL]
[3] [login to view URL]
=== Attachments ===
1_2MyOcr.m = Example script enhanced from [1] based on Coursera Course
[login to view URL] = training letter set
[login to view URL] = example text to detect
Hello there, my name is Daniel and I would love to help you out with this project. I have plenty of experience with Tensorflow for image recognition so I know there will be no problem with this part of the project. I also have experience with Linux and Bash so I can make a bash files that runs everything. Nevertheless, I find it difficult that the main logic of the neural network is no the bash file as it should be on the Python file that is using the tensorflow library. It would be great to discuss further details of your project so I hope to hear back from you soon. Thanks
dear Sir, I can do this project. Im researcher in deep learning, I
can assure you that if you work with me once, you will always work with me for these kinds of projects.