Robotics Software

Ultra-efficient machine learning offers on-device inference powered solely by sound - under the sea

Designed to bring machine learning to the Earth's last great frontier, a prototype smart sensor platform users on-device inference to classify marine mammal calls — powering itself exclusively from the sounds it captures.

Why are location services so crucial for asset tracking?

Logistics companies can manage their supply chains and respond to issues more efficiently when they have a clear and continuous view of the location of assets carried by containers and pallets. Asset tracking enables them to have this. And it goes beyond mere logistics.

Ville Veikko Helppi

 from 

Nordic Semiconductor-

Updated on 10 Mar, 2022

3 minutes read

Internet speed can reach new heights

A completely new type of detector for decoding data transmitted using laser light can help break current speed limits on the Internet.

Lotte Krull

 from 

Technical University of Denmark-

Updated on 10 Mar, 2022

3 minutes read

A Look at What's Next in AI, Deep Learning and Beyond

Developers at all skill levels can learn and grow from attending cutting-edge conferences and leveraging online resources.

The benefits of peripheral vision for machines

Research from MIT suggests that a certain type of robust computer-vision model perceives visual representations similarly to the way humans do using peripheral vision.

Podcast: Wildlife Control & Observation Via UAVs

In this episode, we talk about how UAVs are being deployed to control destructive pigeon populations and to study bottlenose dolphin populations.

Daniel Mitchell

 from 

The Next Byte-

Updated on 08 Mar, 2022

TinyMAN, a reinforcement-learning approach to energy harvesting management, could mean your wearables never need charging again

Designed for "energy neutral operation" of resource-constrained wearable devices, tinyMAN learns how to eke out the power available from an energy harvesting system to maximize utility without ever needing a manual charge.

Transforming wired industrial sensors into IoT devices

The Industrial IoT (IIoT) is the foundation of Industry 4.0. By sending sensor data to the cloud and using it to feed machine learning algorithms, manufacturers can detect process problems immediately, rather than at later stages when rework and field failures are much more expensive.

Lorenzo Amicucci

 from 

Nordic Semiconductor-

Updated on 04 Mar, 2022

3 minutes read

SpeeChin aims to boost privacy by using silent speech recognition for smartphone control

Designed as a necklace-style wearable in a 3D-printed case, SpeeChin is capable of recognizing 54 English and 44 Chinese voice commands — even if the wearer never says anything out loud.

Podcast: AI For Better Battery Design & Training Microrobots

In this episode, we talk about how AI is being leveraged for designing better batteries and training microswimmers.

Farbod Moghaddam

 from 

The Next Byte-

Updated on 01 Mar, 2022

Data Annotation for Computer Vision

5 Questions to Ask Before Getting Started with Data Annotation: What is data annotation? Why is it essential? What are common label types for models? How do AI-powered data annotation tools help with the computer vision labeling process? What’s the best way to get started?

Sage Elliott

 from 

Plainsight-

Updated on 10 Feb, 2022

4 minutes read

Bio-inspired retina-like sensor with visual adaptation

The in-sensor adaptation strategy widens the range for image perception under different illumination conditions to simplify the complexity of hardware and algorithms.

Artificial intelligence listens to the sound of healthy machines

Sounds provide important information about how well a machine is running. ETH researchers have now developed a new machine learning method that automatically detects whether a machine is "healthy" or requires maintenance.

Florian Meyer

 from 

ETH Zurich-

Updated on 28 Feb, 2022

4 minutes read

Virtual Outlier Synthesis Framework for Improving Out-of-Distribution Detection Performance

Designed to address the OOD data false prediction by adaptively synthesizing virtual outliers that can maintain the model’s decision boundary during training.

Physics-aware training turns physical systems, including an oscillating metal plate, into deep neural networks

Designed to address the exploding computational demands of deep neural networks, physical neural networks (PNNs) branch out from electronics into optics and even mechanics to boost performance and efficiency.