Federated learning is a new collaborative learning method that involves multiple decentralized edge devices to build a privacy-preserving shared data model without exchanging locally bound data. This technique is critical to many real-world scenarios where privacy is an issue, such…
In the news | Africa Outlook Magazin
In partnership with Microsoft, the Alliance for a Green Revolution in Africa (AGRA) has launched its new digital tool, the 'AgriBot'. A catalyst for inclusive agricultural transformation, we find out more in a quickfire Q&A with Senior Programme Officer for…
Awards | GSA Technology Transformation Services
Enabling AI and Data-driven Precision Agriculture with FarmBeats was 1 of 4 winners of the GSA’s Technology Transformation Services (TTS) Applied AI Challenge. The challenge was created to improve how the federal government applies emerging AI and machine learning (ML)…
In the news | MSR News
Federated learning is a new collaborative learning method that involves multiple decentralized edge devices to build a privacy-preserving shared data model without exchanging locally bound data.
In the news | IANS
NEW DELHI: Microsoft President and Vice Chair Brad Smith on Thursday visited SEEDS, a disaster response and preparedness non-profit organisation, which is deploying artificial intelligence (AI) to predict the impact of multiple hazards like cyclones, earthquakes or heat waves in…
In the news | Microsoft Stories
Heat waves are becoming more common and frequent around the planet. A non-profit is now deploying AI to help vulnerable communities fight its consequences.
In the news | BBC
Project Origin is part of a wide collaboration between media and tech organisations to develop signals that can be tied to media content to allow audiences to determine where content has come from and check for any manipulation or changes…
In the news | CNBC Squawk Box Asia
Ranveer Chandra is interviewed as part of their “food week” to speak to technology trends in agriculture and how Microsoft is innovating to help industry / growers to sustainably nourish the world.
编者按:近年来,基础模型(foundation models,也被称为预训练模型)的研究从技术层面逐渐趋向于大一统(the big convergence),不同人工智能领域(例如自然语言处理、计算机视觉、语音处理、多模态等)的基础模型从技术上都依赖三个方面:一是 Transformers 成为不同领域和问题的通用神经网络架构和建模方式,二是生成式预训练(generative pre-traini...