#Dog motion capture code#
#Dog motion capture simulator#

They ended up being remarkably dog-like behaviors, despite the robot’s lack of a dog anatomy the researchers even got it to chase its nonexistent tail, spinning around in circles. But some of those behaviors result in a better gait than others.

The system has already made its most catastrophic mistakes in the computer simulation-remember those demerits-so the robot doesn’t have to make them in the real world. “The idea is that if we train the simulation with enough diversity, it might learn a good enough set of strategies, such that one of those strategies will work in the real world,” Peng says.Īll of these strategies are reasonable for the robot to pull off, by the way-they don’t want it to move so rapidly or violently that it will injure itself or humans. They randomized friction in the simulation, for instance, and tweaked the latency between when you send the robot a command and when it actually executes the order. So Peng and his colleagues built not one definitive robot simulation, but a range of possibilities for what the robot’s behavior could be. The new system could be the first steps (sorry) toward robots that learn to move not thanks to exhaustive coding, but by watching videos of animals running and jumping. Not only that, its learned gait is faster than the fastest gait provided by the manufacturer of the robot-though in fairness it’s not yet as stable. But thanks to this translation work, Laikago has learned to move like a real-life canine. (The robot is named, by the way, after Laika, the Soviet space dog who was the first animal to orbit Earth.)Ī robot works quite differently than a biological dog it has motors instead of muscles, and in general it’s a lot stiffer. Then they port those algorithms into the physical version of Laikago. The researchers then translate the digital version of the real dog into a digital version of their four-legged robot-Laikago, which has a rectangular body and skinny legs.

#Dog motion capture simulator#

They gather motion-capture videos from a public dataset, then feed that data into a simulator to create a digital version of the pooch. Over at Google, researchers have a secret weapon to teach robots to move that’s both less taxing and more adorable: dogs. Pity the engineers who have to write out all that code. What comes so easily to you-OK maybe not backflips, just walking-requires extreme coordination, which roboticists have to replicate, a kind of dance of motors working in concert.

#Dog motion capture code#

But what you don’t see is the wildly complex underlying code that makes it possible. As well as recording the colors red, green, and blue (RGB) in every pixel of an image, like a normal digital camera, RGBD cameras also take note of the distance from the camera for each pixel.What you see when Boston Dynamics’ humanoid robot does a backflip or its Spot dog robot fights off a human and opens a door is incredible hardware engineering, to be sure.

This information was then used to train a model that can accurately predict and replicate poses of other dogs imaged by a single RGBD camera. Kearney and her colleagues’ computer model relies on the dataset gathered from some canine friends, including lanky lurchers and squat pugs, who were filmed in their motion-capture suits trotting and jumping. Dog owners could also use it to make a 3D digital representation of their pet on their computer, which is a lot of fun!” “For the entertainment industry, our research can help produce more authentic movement of virtual animals in films and video games. “This technology allows us to study the movement of animals, which is useful for applications such as detecting lameness in a dog and measuring its recovery over time,” Sinéad Kearney, PhD researcher in the university’s motion capture research center (CAMERA), said in a statement. Advertisement Dogs from Bath Cats and Dogs Home were filmed in their suits to create a dataset of movements, that could train a computer model.