Computer Vision

Chrome Extension and Phone workout app using Pose Estimation for Gamification and content Matching

The Chrome Extension utilizes human pose estimation to detect the human and create an image overlay over youtube videos so you can see yourself while you are working out at home. In addition, H10 was contracted to develop a phone workout app that leverages machine learning to better recommend workout videos found on youtube; workouts were based on user preference and videos related to the style and type of instructor.

Tags: Sports Analytics, Computer Vision, Chrome Extension, IOS, Android, App

Multi-spectral Camera System and Machine Learning Image Analysis

A multi-spectral intraoral camera was developed by H10 and other respective collaborators. H10 worked on the optical system and the software development. The software handled both the camera calibration to eliminate lense distortion, the camera control software and the machine learning object detection algorithm. All software was deployed within the camera system.

Tags: Medical, Computer Vision, Multi-Spectral Images

Lacrosse Ball Tracking

Lacrosse is a growing sport in North America but with limited technology to help coaches improve player’s performance. As such, a high-velocity ball tracking system was developed to determine the number and location of each save and goal. This information was development to help improve goalie performance.

Tags: Sports Analytics, Lacrosse, Computer Vision, Object Tracking, Sports

Skin Cancer Detection and Biomarkers Classification

Skin cancer is the most common form of cancer worldwide. To help detect the various kinds of skin cancer, a skin classification algorithm was developed to detect the various types of skin lesions. In addition, a novel biomarker search functionality was created to help identify the differences between each skin lesion type and to provide functionality to compare related lesions.

Tags: Medical, Computer Vision, Cancer Detection, Search

Pose Estimation and Action Recognition in Broadcast Video

Player performance analysis, especially in games, is vitally important in competitive sports such as hockey. To understand player performance human pose estimation and action recognition ML algorithms were used to understand the pose of players (joint locations) and their actions in order to gather actionable insights.

Tags: Sport Analytics, Computer Vision, Hockey, Pose Estimation, Action Recognition, Tracking

Vertical Farm Growth Analysis and Plant Classification

Vertical farms are becoming a more prominent food production source leading to improved horticulture technologies leveraging computer vision and machine learning. H10 designed and implemented technology to track plant growth and classify plants based on the quality and type to understand plant health and plant growing efficacy in all stages of plant growth.

Tags: Horticulture, Computer Vision, Classification