Computer vision and deep learning

Pars Robotics develop and share computer vision and deep learning algorithms. 

“Artificial Intelligence is the new electricity.”

 Andrew Ng

Computer Vision

Humans use their eyes and their brains to see and visually sense the world around them. Computer vision is the science that aims to give a similar, if not better, capability to a machine or computer.

Computer vision is concerned with the automatic extraction, analysis and understanding of useful information from a single image or a sequence of images. 

Image Processing

The second component of Computer Vision is the low-level processing of images. Algorithms are applied to the binary data acquired in the first step to infer low-level information on parts of the image. This type of information is characterized by image edges, point features or segments, for example. They are all the basic geometric elements that build objects in images.

Deep Learning

Deep learning (also known as deep structured learning or hierarchical learning) is part of a broader family of machine learning methods based on learning data representations, as opposed to task-specific algorithms. Learning can be supervised, semi-supervised or unsupervised.[1][2][3]

Why is Deep Learning so popular?

These days, why have we been hearing so much about deep learning? Traditional Machine Learning approaches worked like the top half of the picture above. You would have to design a feature extraction algorithm which generally involved a lot of mathematics (complex design), wasn’t very efficient, and didn’t perform too well at all. After doing all of that you would also have to design a whole classification model to classify your input given the extracted features.

  • With deep networks we can perform feature extraction and classification in one shot, which means we only have to design one model.
  • The availability of large amounts of labelled data as well as GPUs which can process this data in parallel at high speeds enables these models to be much faster than previous methods.
  • Using the back-propagation algorithm, a well-designed loss function, and millions of parameters, these deep networks are able to learn highly complex features (which had to traditionally be hand designed) i.e No more complex design!
  • They’ve become fairly easy to implement, especially with high-level open source libraries such as Keras, Pytorch, and TensorFlow.

What will we do?

We would like to share our experience and teach the complex algortihms to all student!  In the future, we will share instructions for using open source libraries such as Tensorflow and explain deep learning algortihms in a simple way.

We’ll give presentations  about deep learning. In these presentations, we will make an introduction to deep learning. Students will know how to train and predict their models and use NVIDIA Jetson TX1. They will be able to easily install high-level deep learning libraries.

We are using NVIDIA JETSON TX1 to train and predict our models. Our  training videos and papers will be shared!