Machine Intelligence

Auto Added by WPeMatico

Introducing Nested Learning: A new ML paradigm for continual learning

We introduce Nested Learning, a new approach to machine learning that views models as a set of smaller, nested optimization problems, each with its own internal workflow, in order to mitigate or even completely avoid the issue of “catastrophic forgetting”, where learning new tasks sacrifices proficiency on old tasks. The last decade has seen incredible […]

Introducing Nested Learning: A new ML paradigm for continual learning Read More »

The anatomy of a personal health agent

Learn about Google Researchers research prototype, an LLM-powered personal health agent that analyzes data from everyday wellness devices paired with health data, such as blood biomarkers, to offer evidence-based health insights and to provide a personalized coaching experience. The rapid advancement of large language models (LLMs), combined with data from wearable devices, presents a transformative opportunity

The anatomy of a personal health agent Read More »

A collaborative approach to image generation : PASTA

Google Research introduce PASTA, a reinforcement learning agent that refines text-to-image output over multiple turns of interaction with a user by learning their unique preferences. This process is made possible by a novel user simulation technique. You have a perfect image in your mind. You enter a prompt, hit generate, and the result is close

A collaborative approach to image generation : PASTA Read More »

​​Speech-to-Retrieval (S2R): A new approach to voice search

Voice Search is now powered by our new Speech-to-Retrieval engine, which gets answers straight from your spoken query without having to convert it to text first, resulting in a faster, more reliable search for everyone. Voice-based web search has been around a long time and continues to be used by many people, with the underlying

​​Speech-to-Retrieval (S2R): A new approach to voice search Read More »

A picture’s worth a thousand (private) words: Hierarchical generation of coherent synthetic photo albums

We introduce a method for generating differentially private synthetic photo albums that uses an intermediate text representation and produces the albums in a hierarchical fashion. Differential privacy (DP) provides a powerful, mathematically rigorous assurance that sensitive individual information in a dataset remains protected, even when a dataset is used for analysis. Since DP’s inception nearly two decades

A picture’s worth a thousand (private) words: Hierarchical generation of coherent synthetic photo albums Read More »

StreetReaderAI: Towards making street view accessible via context-aware multimodal AI

We introduce StreetReaderAI, a new accessible street view prototype using context-aware, real-time AI and accessible navigation controls. Interactive streetscape tools, available today in every major mapping service, have revolutionized how people virtually navigate and explore the world — from previewing routes and inspecting destinations to remotely visiting world-class tourist locations. But to date, screen readers

StreetReaderAI: Towards making street view accessible via context-aware multimodal AI Read More »

DS-STAR: A state-of-the-art versatile data science agent

DS-STAR is a state-of-the-art data science agent whose versatility is shown by its ability to automate a range of tasks — from statistical analysis to visualization and data wrangling — across various data types, culminating in a top-ranking performance on the famous DABStep benchmark. Data science is a field dedicated to transforming raw data into

DS-STAR: A state-of-the-art versatile data science agent Read More »