This week’s list of top data news highlights covers June 15, 2024 to June 21, 2024 and includes articles on TikTok’s new suite of AI tools for content creators and a digital tool that collects data to better understand migraines.
Researchers at New York-based Cold Spring Harbor Laboratory have developed a new machine learning model that allows individual artificial neurons to receive feedback and adjust in real-time, making AI models more efficient and energy-saving. The new model will allow AI systems to move and process data more efficiently by mimicking how the human brain processes information.
Researchers from Stanford University, University of California, Berkeley, Google DeepMind, and other labs have partnered to release OpenVLA, an open-source vision-language-action (VLA) model for robotics, a type of machine learning that can process both natural language and images and control robots. OpenVLA, which researchers trained using data from real-world robot demonstrations, makes advanced robotics more accessible to a wider range of companies and research labs by allowing developers to fine-tune the model at a low cost for multi-task environments involving multiple objects.
3. Improving Energy Efficiency
Swedish Internet of Things (IoT) network operator Netmore has partnered with a Stockholm-based real estate firm to install low-power wide area network devices in thousands of apartments in Stockholm, Malmö, and Gothenburg. The IoT technology will use data from installed water meters and temperature and humidity sensors to minimize excess consumption and save energy costs for residents, thus also enabling sustainable urban development in those cities.
4. Streamlining Army Acquisitions
The U.S. Army is launching a pilot project that will allow its acquisition and contracting staff to use generative AI to streamline tasks like contract writing and data analysis. The pilot model provides citations that indicate where the data it provides originated, a feature that will help the service fact-check information.
5. Scaling Online Content Creation
TikTok has launched new generative AI tools to help content creators reach global audiences. A new tool called Digital Avatars allows content creators to use AI-generated custom avatars for branding, while a new dubbing feature uses AI to translate and dub content into more than 10 languages.
FranceQCI, a consortium of companies and organizations developing quantum communications, has successfully implemented a quantum communication network using existing fiber optic infrastructure, demonstrating the feasibility of quantum key distribution (QKD), a method to securely share cryptographic keys, on a large scale using existing communications infrastructure. The project combined QKD with post-quantum cryptography to cover an extended distance range by creating a quantum fiber network of around 50 miles that interconnects several quantum nodes in the Paris region.
Researchers at the Hong Kong Polytechnic University in China have created a machine learning algorithm that processes satellite data to more accurately and efficiently predict space weather conditions caused by solar activity. The new model can improve space weather monitoring and improve navigation accuracy in aviation, maritime, and land transportation.
California-based digital health platform Evidation has launched a new tool called MigraineSmart that collects survey and wearable device data on various health metrics, like heart rate, from patients suffering from migraines to better understand the impact of migraines on daily life and help generate new models for migraine treatment. The tool will offer monetary rewards to users who share additional information like pharmacy claims and electronic health records.
Scientists at the University of Dundee in the United Kingdom are developing a new virtual reality (VR) technology to assist investigators in determining the causes of fires and preventing future fire hazards. The new technology uses real-world footage and turns it into a VR-powered immersive representation of fire scenes, which investigators then study during their training.
10. Generating Video Soundtracks
Google DeepMind has unveiled an AI tool that generates synced audio and video soundtracks using text prompts and video content analysis. This tool can create various audio effects, such as drama scores or realistic sound effects, synced with videos and offers an unlimited number of soundtrack options.
Image credit: Jay Heike
