An analysis of a Vanishing Point Detection architecture employing a Heatmap-Based Regression to estimate Vanishing Point likelihoods. Research on Vanishing Point (VP) Detection has become of more and more interest in recent years, with several proposed approaches either building on previous work or exploiting new techniques. Autonomous Navigation and Autonomous…

A brief historical review of the DeepLab segmentation architecture: from DeepLabv1 to DeepLabv3+. Image segmentation tasks have seen lots of developments in recent years, and have become one of the most researched topics in Computer Vision⁶. One of the standards for segmentation is represented by the Deep Labelling for Image…

A discussion of one of the most popular architectures for image segmentation: from Fully Convolutional Networks to U-Net and V-Net. Image Segmentation is a fundamental task for Computer Vision, dividing a scene into meaningful regions/segments to achieve a better understanding and processing of an image. Segmentation tasks have been at…

A discussion of two state-of-the-art architectures for real-time Vanishing Point Detection using limited computational resources. Vanishing Point Detection is a fundamental task in visual positioning systems, with lots of applications for Autonomous Driving and Autonomous Navigation⁶. While the performances of new approaches keep increasing, maintaining the complexity of such systems…

An introduction to the concept of Vanishing Point, its importance for Computer Vision, and intuitions behind the most popular approaches for detection. Computer Vision, Autonomous Driving, and Autonomous Navigation, among other fields, have become increasingly popular and a multitude of applications employing such methods are already in place or are…