Sleip releases upgraded AI

Sleip releases upgraded AI

Sleip now measures over 100 anatomical key points on the moving horse

Sleip is proud to announce a significant update in our AI capabilities, enhancing user experiences and speeding up analysis. The upgraded neural network also opens the door for exciting product development opportunities.

At the heart of Sleip's technology is the innovative use of computer vision, a specialised sub-field of artificial intelligence (AI) that mimics human visual understanding – but far surpasses our cognitive capabilities. The new release catapults Sleip’s AI from about 20-30 to over 100 anatomical measurements in each frame of video.

 

Our AI now tracks 115 specific points on the horse's body in each frame of video footage, meaning roughly 350,000 measurements for an average trot up. These points act as markers, much like sensors attached to horses in gait analysis systems that employ physical equipment. And with this, we can measure movements rapidly in every video frame, says Felix Lawin, Computer Vision Researcher at Sleip.

Felix Järemo Lawin

PhD, Computer Vision Researcher

felix

 

Vertical displacement of head and pelvis to measure asymmetries

Based on decades of equine biomechanical research, the gold standard for gait analysis that quantifies asymmetries is the measurement of the vertical displacement of the head and pelvis during a trot. However, by capturing a wider array of points, we get a more comprehensive view of a horse's movement pattern. This ensures the most accurate strides are chosen for analysis.

Enhanced tracking capability makes recordings easier 

This comprehensive tracking allows Sleip to select the most optimal strides for gait analysis measurements. It ensures that our system can accurately understand and measure the horse's movement, even if there are some obstructions in the camera view, such as  fencing when recording longeing in a round pen.

This next-generation neural network also sets the stage for future opportunities. With more detailed and holistic motion measurements, we can extract more data and detect more patterns from Sleip recordings. The potential to bring valuable insights for equine health and performance management is considerable, concludes Felix Lawin.

Latest news

Create Sleip videos at walk and canter

Create Sleip videos at walk and canter

Sleip now automatically identifies horse movement as walk, trot, or canter.

Sleip listed as one of the “most promising AI-startups”

Sleip listed as one of the “most promising AI-startups”

Sleip listed as one of the “most promising AI-startups”

Stride frequency in the Sleip app

Stride frequency in the Sleip app

We are pleased to release a new basic biomechanical measurement! Sleip now includes data on the horse’s stride frequency in the analysis summary, measured in strides per second.