DOMO City Search
Re-discovering the city in Temporal-spatial patter
We developed a LGFnet Deep Neural Network Model, the can easily recognise the building at every condition, not matter at what angle, no matter under what weather, and no matter in what time. All information in your photo will be extracted by comparing it with city database. You can even get more: route planning, attractions guide, restaurant recommendation, business assistant, or just simply let you know where you exactly are.
This project began in January 2018 and was selected as one of the final Top 5 project of the Copenhagen AI Hackathon 2018. Other teams come from the UK, France, Singapore, Denmark and Sweden, including Sony and Capgemini. Danish national television (TV 2) reported the news. read more
DOMO Animal Health
Knows Animal More
We are aiming to develop IOA (Internet of Animals) edge-computing product for early detection and early alarms of animal health status, at easy implementation and low maintenance cost. On such edge-computing device, our multiple-stage Deep Neural Network model (A-Idea Model) is running. A-Idea model can track and distinguish individual animal, detect fine movement of animal, and identify tiny details of animal behaviors. Animal health status detection can be thus made at very early stage and done locally at edge devices, rather than transferring sensor data to data center (or high-performance computer). Only important alarm message is sent over internet. In such way, required data flow over internet is low, making practical implementation easy and maintenance cost low.
We completed the first prototype product in August 2019, follow the progress of this project, please visit Mr. BONDE, an AI Camera Kit.
This project is the winner of Nordic/China Hi-Tech Weeks 2019, selected from a wide range of startups in Copenhagen, Oslo, Stockholm, Uppsala-Lund. And also help us win the Best University Startup, Only one representative from Sweden, University Startups & Spin-Offs Festival 2019.