EUROPEAN MACHINE VISION FORUM
Bonseyes EU project to bring AI to embedded devices
An €8.5 million European project to bring artificial intelligence development to low-power cloud and edge devices has reached the midway point in its three-year duration. Professor Oscar Deniz Suárez, from the Visilab
group of the Universidad de Castilla-La Mancha, in Ciudad Real, Spain, described the Bonseyes project, along with the Eyes of Tings project, which has just finished, in a keynote presentation at the European Machine Vision Forum in Bologna, Italy. Bonseyes aims to be an AI marketplace, a
repository where labelled data and trained models are shared. Te site will also contain a deep learning toolbox to help users create their final model for deep learning inference, a model that will be able to run on an embedded device. Suárez said that intelligence on the edge – or
onboard a device – is not only advisable in terms of user experience, it is also probably mandatory because there’s not enough bandwidth to stream images to central servers and respond to that data. Tere is a trend for edge processing, according
to Suárez, and to have many small devices able to run cognitive data processing and therefore provide instant response to the user everywhere. Tere are, however, some obstacles to this, a big one being that to do something useful with deep
learning requires labelled data and lots of it. Tis is what Suárez refers to as the ‘data wall’, which is what the Bonseyes project hopes to remove to some extent. ‘If you use the cloud, then you shouldn’t have
any problem with deep learning,’ Suárez said. ‘Te problems start to come when you want to use [deep learning] in an embedded device – in a car, a robot, a toy – especially in terms of power consumption, latency, memory.’ It’s these challenges that Bonseyes is looking to tackle. Te project’s deep learning toolbox should help users carry out deep learning inference on the edge. Te marketplace’s front-end will be a web- based GUI, while the AI artefact layer, the
back-end, will be made up of a typical machine learning pipeline. One of the steps in the pipeline will convert a model into something smaller that will run on an embedded device. Te project partners, which include big names
such as Arm, are currently investigating methods that would approximate a model into something more lightweight, so it takes less memory, and can run with lower latency and lower power consumption. Bonseyes tools and programming
environments will be supported by several reference platforms with pre-integrated middleware via open source and vendor packages. Suárez also spoke about the Eyes of Tings
project, which is an open hardware platform based on Movidius’s – now Intel – Myriad2 system-on-chip. Te prototype board presented at the project’s final meeting cost €112 for the bill of materials, although this was for a prototype volume, and Suárez said it could be made cheaper. As part of the project, the partners made an
automatic museum audio guide, building a device with a camera attached that the user would wear. Te demonstrator was programmed to recognise paintings and relay information about the artist to the wearer.
Robot uses vision to infer human actions to work together better
Engineers at the Politecnico di Milano, in Italy, are developing a vision system that can infer what a human is going to do in order for a robot to operate more effectively with its human co-workers. Professor Paolo Rocco at the
Politecnico di Milano gave a presentation at the European Machine Vision Forum in Bologna, in September, about the research his group is conducting on cobots, robots designed to work with humans. One of the projects he spoke about is a vision system that can predict a human’s intention by looking at how they’re working. In
an assembly process, for instance, once the robot knows what part the human will pick up, it can move to another task or aid the human. Te system uses a deep learning
algorithm trained, in the example Rocco showed, by tracking the motion of the human’s right hand. In this example the engineer was assembling a box by adding a lid to it. Te human and robot are in front of one another and between them are two trays:, one with the box, the other with the lid. Whenever the human goes to take a box or pick up a lid, the robot understands and does the complementary action. So,
16 Imaging and Machine Vision Europe • October/November 2018
when the human takes the lid, the robot picks up the box and positions it for the human to attach the lid, and vice versa. ‘Humans normally do a sequence
of actions,’ Rocco explained. ‘Te most recent thing we’ve done is to predict the sequence of actions the human will do. So, whenever the human wants to make a collaborative action, the robot will be ready, so there will be no waste of time.’ In Rocco’s setup, the robot
constantly monitors the human’s action so that the cycle of the human and that of the robot
synchronises. ‘Te benefit was a 20 per cent reduction in cycle time just using the machine learning algorithm to infer what the human is doing,’ Rocco said. Rocco’s group has recently
founded a spin-off firm called Smart Robots. It makes a smart hardware and soſtware solution, which optimises the space where human operators and robots work side by side. Te device can be linked to robots from various manufacturers, and can be used to ensure collision avoidance, or reschedule the robot’s tasks if it interferes with what the human is doing.
@imveurope
www.imveurope.com
Metamorworks/
Shutterstock.com
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60 |
Page 61 |
Page 62 |
Page 63 |
Page 64 |
Page 65 |
Page 66 |
Page 67 |
Page 68 |
Page 69 |
Page 70 |
Page 71 |
Page 72
