Object detection

Object detection is a computer vision/image processing task where we need to identify instances of semantic objects of a certain class in an image or video. In classification problems like those discussed in the previous section, we simply need to assign a class to each image, whereas in object detection tasks, we want to draw a bounding box around an object of interest to locate it within an image.

In this section, we will use data from the Global Wheat Detection competition (https://www.kaggle.com/c/global-wheat-detection). In this competition, participants had to detect wheat heads, which are spikes atop plants containing grain. Detection of these in plant images is used to estimate the size and density of wheat heads across crop varieties. We will demonstrate how to train a model for solving this using Yolov5, a well-established model in object detection, and state-of-the-art until late 2021 when it was (based on preliminary results) surpassed by the YoloX...