Sign in

AI Engineer | Deep Learning | Artist

Simple tips and tricks to increase FPS without custom training model.

Photo by Matthew Brodeur on Unsplash

What is Detectron2?

Detectron2 is a next generation software system developed by Facebook AI Research for Object Detection.

Object detection means to recognize, localize and predict attributes of objects in an image.

You can find the repository here .

Detectron2 is a repository that can be used for detection tasks such as instance segmentation, bounding box detection, person key point detection and semantic segmentation.

The datasets used are COCO(Common Object in Context) , LVIS(Large Vocabulary Instance Segmentation) , CityScapes, PascalVOC.

Detectron2 is already fast and inference time is less. …


Why the kernel size should be odd? What happens if we use even kernel size?

Photo by R Mo on Unsplash

Before we start with significance of kernel size lets first revisit what kernel really is.

In Fully-connected network every vector element in the input is connected to every hidden unit in first layer. So you connect every unit to every other unit.

So if you have 11x11 image then there would be 121 connections to a single unit and in practice there would be 363 connections. Since every image has 3 channels.

In locally-connected network we use a kernel/filter. Kernel is a part of image…


Photo by Backpack Studio on Unsplash

What is Edge computing ?

Edge computing is computing that’s done at or near the source of the data, instead of relying on the cloud at one of data centers to do all the work.

It doesn’t mean that the cloud will disappear. It means the cloud is coming to you.

Why do we need edge computing ?

Edge computing brings the data and computation closer to where it is actually needed. Another alternative used these days is cloud computing which is not cost effective and there might be latency. Also there is always security factors.

NVIDIA Xavier is an embedded computing board from NVIDIA Jetson series.

  • CPU: Nvidia custom Carmel ARMv8.2-A…


Photo by Joshua Sortino on Unsplash

In Data Science, the most important aspect is gathering data. We may get data through various platforms which includes Kaggle, Google’s dataset, Visual data etc.

However if you want to collect data from a particular website like Medium which are auto loading pages (i.e infinite scrolling page) we need to use web scraping.

In this post we will learn how to web-scrape data using Beautiful soup. For infinitely scrolling pages for which we will use selenium.

Selenium is a web framework that permits you to execute cross-browser tests. Through Selenium Python API you can access all functionalities of Selenium Web…


Curvature detection without using Neural Network

Photo by Raphael Schaller on Unsplash

We are familiar with the term Autonomous cars. Where the car manages to navigate from one place to another without human assistance or intervention using a complex pipeline that includes “artificial neural networks”. However, is this possible without Neural Networks?

While navigating our route, we may face different types of scenarios like curved roads where the car needs to control its steering and brakes based on the decisions made by its Neural Network. In this post we will see how image processing can be useful to determine in real-time if the car is supposed to go straight or take turn.

Anuja Ihare

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store