|

Here is a little background

Currently, I am working at American Wild Horse Campaign, where I am trying to learn how to use deep learning models to detect horses in images. My work consists mainly of creating models for object segmentation, object classification and quality assessment of images for the horses. The goal of my work is to create a filter for an app that the company is going to use to collect large amounts of data for training a model that will be able to recognize horses. Unlike humans, who have a specific marker like fingerprint or faces, horses can look practically identical, so we choose to use angles, stance and poses to segment them. This work is important for the activists in the company who go out into the wild to dart horses and track horse populations and fertility - they need to know how many they have gotten etc.