For last week’s blog post, I wrote a short tutorial for training a custom object detection model using TensorFlow Object Detection API. Due to the limited space and time constraints, my tutorial was not quite finished. Therefore, in this week’s blog, I will continue my tutorial and include additional steps such as the usage of a tool to test your model’s accuracy.
Have you ever heard of Tesla’s Model S sedan? It is one of the few cars capable of fully autonomous driving. Although U.S. laws currently do not permit this, the Model S can pick you up at your house and drop you off at school, all without you even touching the steering wheel. To create a self-driving vehicle, Tesla engineers had to employ many machine learning techniques, including an object detector that recognizes and classifies objects around the car. For example, the on-board camera is able to recognize pedestrians and instructs the car to stop. Another example is that the object detector recognizes other vehicles on the road, keeping the Tesla from colliding into them.
With the use of the TensorFlow Object Detection API, creating such a model (though probably not as accurate as the one Tesla developed) can now be done with consumer-grade equipment such as a personal computer. As promised in last week’s blog, I will discuss how to create a customized object detector with the TensorFlow API.
In last week’s blog, I wrote about the installation process of TensorFlow on a PC. In this blog, I will continue my tutorial on TensorFlow installation.
Step 3: Installing TensorFlow
At the time of writing, there are two supported ways of installing TensorFlow:
- Native pip
While native pip installs TensorFlow directly onto your computer, Anaconda allows you to create a virtual environment and install TensorFlow into that environment. The benefit of this is to help you avoid unwanted interference with other packages. However, if you do chose to use Anaconda, you will not be able to access the TensorFlow package globally (from any directory on your computer). The following chart compares the two different installation methods:
Instead of writing a weekly update on my project, I would like to take a moment to reflect on what I have been able to achieve in Quarter 3. I will also discuss the plans I have for Q4 as well as my final product. Continue reading
It has been quite a while since I last wrote about machine learning. During the first week after spring break, I finally got a chance to take a closer look at TensorFlow, a software library designed for machine learning applications by Google. It was originally used by the Google Brain team and was later made open-source to the public. In this week’s blog post, I will discuss the installation of TensorFlow on PCs.
During this past week, I devoted the majority of my time to working on the next Polaris major release (codename “NX”). Among the numerous features that will be made available, “remote printing” is the most requested. It will allow duty crew members to print attendance sheets directly through the Polaris cloud printing service, eliminating the need for the driver take attendance by hand with the van sign-out form.
I am generally not a fan of any video games. Ubisoft’s Watch Dogs series is an exception. During the long weekend, I had the opportunity to replay Watch Dogs 2 and ended up finishing the all the missions in the story. You might want to ask: how is it game even relevant to your project? The answer is that after a semester’s research on artificial intelligence and big data, I now have a much deeper understanding of the plot of the game.
WEST CHESTER, Pennsylvania — January 31, 2018 — Today, The Polaris Team released the Polaris February Update. The primary goal of this update is to improve the user experience of the administrators and weekend duty crews.
WEST CHESTER, Pennsylvania — January 31, 2018 — Today, Kevin Wang, developer of Argus and Count: A Very Simple Counter (Count), released Count 3.0 to the general public on the App Store. Originally released in 2015, Count is a simplistic and intelligent counter that allows users to count with gestures. The new Count has been completely rebuilt from the ground up and comes with new features including an improved UI, Sense, Random, and various bug fixes and stability improvements.
Over last semester, I wrote a series of blogs focusing on artificial intelligence and machine learning using artificial intelligence, especially that of image classifications. My independent project aimed to create and improve a convolutional neural network that identifies different categories of grocery. Through the semester, I gained considerable experiences working with Tensorflow, the most popular programming framework for machine learning. I also became proficient in creating and improving the neural network, raising its accuracy over 80 percent. Continue reading