This week’s list of top data news highlights covers November 11, 2023 to November 17, 2023 and includes articles on trying on high fashion in augmented reality and training AI systems to classify bird sounds.
1. Forecasting the Weather
Google DeepMind has created an AI system that can create weather forecasts faster, more accurately, and with less computing resources than current models. The team trained the system on nearly 40 years worth of historical weather data.
2. Editing Genes
Researchers at Verve Therapeutics, a U.S.-based biotechnology company, have created a gene-editing treatment for cholesterol. It uses CRISPR editing techniques to disable certain liver gene cells which will help lower bad cholesterol.
3. Packing Online Orders
Amazon has unveiled a robot that can identify the shape of online orders and pack them in an appropriately-sized box. By using the robot, the company can prepare orders for shipment more efficiently and reduce unnecessary packaging materials.
4. Enhancing Work Meetings
Microsoft has updated its Teams application to include two new AI-powered features, including one that can remove clutter from users’ video backgrounds or generate decorations for them, and another that can isolate users’ voices from background noise during calls.
5. Inspecting Plumbing
The Louisville Water Company is using two robots to inspect large underground water pipes in Louisville, Kentucky. The first robot uses a camera and ultrasonic sensors to determine the conditions of pipes and pinpoint potential repairs while the second uses acoustic sensors to identify leaks or air pockets.
6. Composing Music
Google DeepMind has partnered with YouTube to launch a generative AI system for creating music. They have also developed two new tools built on Lyria, including one that can create music in the style of select popular artists for YouTube Shorts and another that can generate instrumental music from a hummed tune, simple keyboard tune, or another vocal line.
7. Editing Videos
Meta has launched two new AI-powered video editing tools. The first can generate four-second videos from a caption or photo and the second can edit videos according to users’ text prompts.
8. Filtering Sounds
Adobe is developing a tool that uses an AI system to separate sounds in an audio recording. Users can upload a recording and select sounds to filter out or edit, such as applause, speech, traffic, and more.
9. Classifying Birds
Researchers at the University of Moncton in Canada have created an AI system that can generate realistic bird songs to train future AI systems to recognize bird songs in the wild. The team used over 23,000 recordings of 264 bird species around the world. In tests, the team found that adding generated bird songs to AI systems for bird identification improved classification by 12 percent.
10. Exploring High Fashion
The Fine Arts Museums of San Francisco has partnered with Snapchat to add mirrors at the de Young Museum. Visitors can use the mirrors to virtually try on three evening outfits from the museum’s high fashion exhibit in augmented reality.