Closed. This question needs to be more focused. It is not currently accepting answers.
Want to improve this question? Update the question so it focuses on one problem only by editing this post.
Closed 3 years ago.
Improve this question
I have hundreds of images of handwritten notes. They were written from different people but they are in sequence so you know that for example person1 wrote img1.jpg -> img100.jpg. The style of handwriting varies a lot from person to person but there are parts of the notes which are always fixed, I imagine that could help an algorithm (it helps me!).
I tried tesseract and it failed pretty bad at recognizing the text. I'm thinking since each person has like 100 images is there an algorithm I can train by feeding it a small number of examples, like 5 or less and it can learn from that? Or would it not be enough data? From searching around it seems looks like I need to implement a CNN (e.g. this paper).
My knowledge of ai is limited though, is this something that I could still do using a library and some studying? If so, what should I do going forward?
This is called OCR and there has been a progress. Actually, here is an example of how simple it is to parse an image file to text using tesseract:
try:
from PIL import Image
except ImportError:
import Image
import pytesseract
def ocr_core(file):
text = pytesseract.image_to_string(file)
return text
print(ocr_core('sample.png'))
BUT
I am not very sure that it can recognize different types of handwriting. You can give it a try yourself to find out. If you want to try the python example you need to import tesseract but first things first to install tesseract on your OS and add it to your PATH.
There are many OCRs out there and some perform better than others. However, this is a field that has improved a lot recently with the Deep Neural Networks. I would consider using a Cloud provider such as Azure, Google Cloud or Amazon. Your upload the image and they return the metadata.
For instance:
https://azure.microsoft.com/en-us/services/cognitive-services/computer-vision/
If you don't want to use cloud services for any reason, I would consider using TensorFlow... but some knowledge is required:
Tensorflow model for OCR
Related
Closed. This question needs details or clarity. It is not currently accepting answers.
Want to improve this question? Add details and clarify the problem by editing this post.
Closed 2 years ago.
Improve this question
I’m learning Java and have recently started my first project. The idea of this project is to pass one input argument - path to file/folder, which would be analyzed in order to find all files with predefined extension, parse them and create objects based on the results of parsing to store for future.
So far I’ve written all the code and my project structure (simplified) looks like that:
Class defining resulting object
Class that analyzes the input parameter (exists, is file, is folder) and processes it, returning list of all suitable files
Class that parses suitable files and creates objects
The question is - am I following OOP with that structure?
From what I’ve read on the web the last two classes seem to look like polterheists. But I don’t think that it is a good idea to move the logic of the third class to the object class because it consists of lots of methods (define current section of the file, strategy to parse each separate section).
I am learning on my own and don’t want to start my journey by cultivating bad habits.
I am learning on my own and don’t want to start my journey by cultivating bad habits.
You're saying this like you have a choice :)
From what you described it seems reasonable, of course w/o seeing the code we can't say. And even if you show the code - 100 people will have 100 opinions, there's a lot of debates around OOP.
What's important is not to look at your design as something static. Once your app starts to be more complicated you'll have to re-work some of it.
PS: stackoverflow doesn't like this kind of questions since everyone will have an opinion. You'll have to find other resources if you keep having such questions.
Closed. This question needs to be more focused. It is not currently accepting answers.
Want to improve this question? Update the question so it focuses on one problem only by editing this post.
Closed 3 years ago.
Improve this question
I used Android Speech-to-Text API (Recognizer Intent), for recognition of the word said by the user. But the problem is that it returns the accurate word after autocorrection. I want it returns the exact word (without correction) said by the user. Please suggest me any other android library for this feature or how can I got my feature inside the android inbuild speech to text API.
I also saw the google API for this but that is paid And that is also AI-based.
I want it to return the exact word (without correction) said by the user.
I think you misunderstand what speech recognition is capable of doing.
A speech recognizing system is only capable of recognizing an uttered word as being one of a number of possible words. It doesn't ... and cannot ... tell you with 100% accuracy what the speaker actually said.
This applies to any speech recognition system, including a human listener. (How many times have you had to ask someone to "Say that again please" ?)
The only way to determine with absolute certainly the exact words that were spoken is to ask the person who spoke them to type them in! (And even then, they may not give you a 100% accurate answer, in some cases.)
In short, what you want is not possible. Software cannot do it. Humans cannot do it, even if they believe that they can1. You need to adjust your expectations.
1 - The Two Ronnies - Four Candles sketch
Identifying / recommending better (more accurate) speech recognition software or services is off-topic.
Closed. This question needs to be more focused. It is not currently accepting answers.
Want to improve this question? Update the question so it focuses on one problem only by editing this post.
Closed 6 years ago.
Improve this question
I am new to the image processing discipline, currently I am working on an activity to find the path in a given floor plan image using java. I could understand & implement a plain Dijkstra or A star algorithm in java to find the shortest path between nodes that can work on hard coded values for nodes & edges. But I have no idea on how to do the same with image file as input.
I could not find any much detailed or comprehensive solution for this on web surfing. Any idea on how to achieve this. Kindly help.
If the walls on your plan are black, and the space is white, then use the white pixels as the space in which you can search using your algorithms.
be aware that:
The plan images could need some preprocessing, like adjusting brightness/contrast and/or converting into lower resolution picture.
If the plans are more complicated than that, ie. you need to take into account some special signs or structures on them then you need to do more complicated conversion/preprocessing and/or image recognition, using some special tools, maybe even some AI.
Closed. This question needs to be more focused. It is not currently accepting answers.
Want to improve this question? Update the question so it focuses on one problem only by editing this post.
Closed 9 years ago.
Improve this question
I am looking into finding algorithm within the area of clustering or machine learning which will facilitate or creating a typical data reading for a group of readings. The issue is that it must facilitate time series data; thus some traditional (k-means) techniques are not as useful.
Can anyone recommend places to look or particular algorithms that would provide a typical reading and relatively simple to implement (in Java), manipulate and understand?
As an idea. Try to convert all data types into time, then you will have vectors of the same type (time), then any clustering strategy will work fine.
By converting to time I actually mean that any measurement or data type we know about has a time in its nature. Time is not a 4-th dimension, as many think! Time is actually 0-dimension. Even a point of no physical dimensions which may not exist in space, exists in time.
Distance, weight, temperature, pressure, directions, speed... all measures we do can be converted into certain functions of time.
I have tried this approach on several projects and it payed back with really nice solutions.
Hope, this might help you here as well.
For most machine learning problems in Java, weka usually works pretty well.
See, for example: http://facweb.cs.depaul.edu/mobasher/classes/ect584/weka/k-means.html
Closed. This question needs to be more focused. It is not currently accepting answers.
Want to improve this question? Update the question so it focuses on one problem only by editing this post.
Closed 3 years ago.
Improve this question
I want to write a code either in Java or PHP (Codeigniter) to extract information such as email and phone number of a user uploading hbis resume or cv to the site. Basically I want to build a cv parser.
Need help for this.
thanks
EDIT
The cv format will be in doc.
Since there is no standard CV format, parsing will be next to impossible.
Instead, consider collecting contact information in an HTML form when they upload.
I'd suggest you to build it using a set of regular expressions.
If you just want to extract phone number and email the parser is very simple. It will work almost 100% for emails and (I believe) 98% for phone numbers.
If you wish to extract other information it will be more complicated because there is no standards for CVs; information may be formatted using different ways. Anyway, good luck!
you should use python and write your own scraper, its easy and it can be done really quickly in your case with modules like beautiful soup, urllib2 ...
what its this all about
beautiful soup documentation
Ditto AlexR. If ALL you want to find is email address and phone number, you could scan for strings of characters in the appropriate format. A couple of simple regular expressions could do that fairly reliably. Even that wouldn't be 100%. If someone included, "Learned Java#Technocorp. US citizen." etc, you might easily be fooled into thinking that's an email address of "java#technocorp.us". Okay, that's a strained example, but it's the sort of thing that shoots down natural language parsing.
If you want more than that, there is no easy answer. You could search for keywords, like to find where he went to school you could look for the words "college" or "university". But even then, someone might put "Graduate of Foobar College" or "College: Foobar" or "BA from Foobar" or many many other possible formats.
As #Corbin said, there is no standard CV format. It will be quite difficult to parse with 100% accuracy.
Though, you can try Apache Tika - A Content Analysis Toolkit to parse resume doc/docx format. Apache also support many document format including pdf, txt, xml, odf etc.
Btw, extracting email and phone number from resume can be achieved with few lines of code with the help of regex after getting whole contents from cv using Apache Tika.
Let me know if you get stuck.
Hope this helps!
Note- (I am working on resume summarizer).