– DROWNING DETECTION SYSTEM
– IOS/ANDROID POSE ESTIMATION/OBJECT DETEION SDK
Underwater Camera > If drowning > Alarm
Swimming is a hotspot in many places. Our Smart Swim solution can address a great demand by the swimmers’ community: Safety. There are many accidents happening in the swimming pools. According to the Leisure and Cultural Service Department (LCSD), there are 328 accidents and 2 drowning deaths in 2018. Our Smart Swim – drowning detection system is being demanded by the market.
Although there are a few solutions for drowning detection exist in the market. However, the solution was developed based on an outdated software, databases and data analysis model. The system can only be used for drown detection as a safety facility only. Moreover the cost is very expensive. It costs around HK$1 million to 2 million for a standard swimming pool installation. We target to develop a more powerful and faster solution at a lower cost so as to make this solution more affordable and could be commonly implemented in the school community.
This solution was one of the deliverables from the ITF Better Living Fund project and the prototype was successfully installed in Munsang College in Sept 2020. It is now getting into the commercialization stage and targets to be deployed in more schools and promoted to the LCSD public pool in the coming year.
Demonstration on Pose Estimation
Underwater with LSTM Action Prediction
Drown Detection Using Computer Vision
Our company focuses on developing technologies related to AI and computer vision. Our three core technologies include object detection, pose estimation and human activity recognition. Previously we have been using our technologies in an AI server, which includes a lot of GPU. However, It is inconvenient for the users to use our technologies in mobile or embedded devices. Actually we have been doing research and development on sliming our model, which we have put our model in mobile phone like IOS and Android device.