Tensorflow - Computer Vision

Popular posts from this blog

Django middleware - Crawler detection

Sometimes you have to detect search engine crawling activity on your system, to handle the workflow for this type of request. On our website, there is a high crawling activity, because BV FAPESP (https://bv.fapesp.br) provides useful information abour reaserch for science, technology and academy in Sao Paulo state of Brazil. Recently, I deployed a feature on our system to allow people to store their queries and filters they do on our website. This is an open feature for everybody who navigates on our system. It is based on the HTTP session of the brower, which is stored on Python/Django server-side. When we were tunning Django session table in the database, we found out that there were a lot of sessions created by crawling activities. To take control of the session creation, I wrote the middleware that follows.  The list of crawlers, I get from ngix logs, on a very small timeframe, so maybe there are some more search engines that are not listed. The use-case presented, ...
Read more

Atom - Jupyter / Hydrogen

Image
I will show you how to use the Atom editor with Jupyter Notebook to debug Python/Django views using Hydrogen plugin. On the Internet you can find some examples using this setup to debug Python but I found nothing related to Django, so I decided to publish this notes. This is not a high performance setup, because it involves many technologies at several layers. This is more a proof of concept of nowadays technologies which are available to the programmer's toolbox. This notes has been written during the development and improvement of the automated tests of the Virtual Library of FAPESP ( http://www.bv.fapesp.br ). The environment The diagram below shows an actual infrastructure to use this setup in a development environment. Technologies It is not expected that you know all of the technologies in detail, but there are some trick points. This paper is oriented for Python programmer who already use Atom and Jupyter and enjoy explore the edges of these tools. ...
Read more

SOLID - Training models in TensorFlow

SOLID - Training models in TensorFlow The SOLID principles can be applied to the software architecture surrounding the machine learning models. Here's how the principles can be relevant: Single Responsibility Principle (SRP) Each class or module in your TensorFlow project should have a clear and single responsibility. For example, you can have separate modules for data preprocessing, model training, model evaluation, and model deployment. This promotes modularity and makes it easier to understand, test, and maintain each component. Open-Closed Principle (OCP) By designing your TensorFlow project with the OCP in mind, you can make it easier to extend the functionality without modifying existing code. For example, you can define abstract base classes or interfaces that define the common behavior expected from different models, allowing you to add new models by implementing these interfaces without modifying the existing code that consumes them. Liskov Substitution Principle (LS...
Read more