\n<\/p>\n
In the eight years since we established our FAIR hub in Paris, Meta has become one of the leading research organizations in the world, with pioneering work stemming from our tech hubs in Paris, London, Tel Aviv, and Zurich.<\/span><\/p>\n One of the most important decisions we made when we set up FAIR was to put exploratory research and open science at the center. We regularly collaborate with external researchers, because we have a strong hypothesis that this is the fastest and most responsible way to make progress.<\/span><\/p>\n \u201cWe have worked with institutions to develop generations of AI researchers, especially via our PhD programs,\u201d said Naila Murray, head of FAIR EMEA. \u201cMany of our PhD students have made important contributions to the field.\u201d<\/span><\/p>\n<\/blockquote>\n Today, our teams in Paris, London, Tel Aviv, and Zurich, are focused on a variety of interests, including self-supervised learning, reinforcement learning, speech and audio, computer vision, natural language modeling, responsible AI, machine learning theory, model efficiency, AR\/VR, and more.<\/span><\/p>\n \u201cOur research is driven by a unique mix of ambition and collegiality, and our team works tightly together across boundaries of expertise, seniority, location, and job role to make rapid research progress,\u201d Murray said. \u201cIn this current era in AI research, seemingly each day brings a potential new research breakthrough, including from our EMEA team.\u201d<\/span><\/p>\n<\/blockquote>\n Earlier this year, our researchers in Paris formed the team that built and deployed<\/span> LLaMA<\/span><\/a> (Large Language Model Meta AI) \u2013 a state-of-the-art foundational <\/span>large language model<\/span><\/a> designed to help researchers advance their work in this subfield of AI.\u00a0<\/span><\/p>\n LLaMA works by taking a sequence of words as an input and predicts a next word to recursively generate text. To train our model, we chose text from the 20 languages with the most speakers, focusing on those with Latin and Cyrillic alphabets. <\/span>With capabilities to generate creative text, <\/span>solve mathematical theorems<\/span><\/a>, <\/span>predict protein structures<\/span><\/a>, answer reading comprehension questions, and more, large language models are one of the clearest cases of the substantial potential benefits AI can offer at scale to billions of people.<\/span><\/p>\n Also based in Paris, our teams introduced two breakthroughs in computer vision research. In April, we unveiled <\/span>DINOv2<\/span><\/a> \u2013 the first method for training computer vision models that uses self-supervised learning to achieve results that match or surpass the standard approach used in the field.\u00a0<\/span><\/p>\n DINOv2 can discover and segment objects in an image or a video with absolutely no supervision and without being given a targeted objective. For example, DINO can understand that an image contains a representation of a dog without ever being taught what a dog is in the first place. As part of this announcement, we shared a <\/span>public demo <\/span><\/a>that anyone can use to explore some of the capabilities of DINOv2.<\/span><\/p>\n We\u2019re already using DINOv2 to learn more about the physical world. Meta recently collaborated with the <\/span>World Resources Institute<\/span><\/a> to <\/span>use AI to map forests<\/span><\/a> \u2013 tree by tree \u2013 across areas the size of continents.<\/span> While our self-supervised model was trained on data from forests in <\/span>North America, evaluations confirm that it generalizes well and delivers accurate maps in other locations around the world.<\/span><\/p>\n Our Paris team, in collaboration with colleagues in North America,\u00a0 also pioneered new research using <\/span>\u00a0SEER (SElf-SupERvised), Meta AI Research\u2019s groundbreaking self-supervised computer vision model<\/span><\/a>. SEER learns directly from any random collection of images \u2014 without the need for careful data curation and labeling that goes into conventional computer vision training \u2014 and then outputs an image embedding.\u00a0<\/span><\/p>\n For our latest breakthrough, SEER10B, we use diverse datasets to enable better and fairer computer vision. <\/span>Traditional computer vision systems are trained primarily on examples from the U.S. and wealthy countries in Europe, so they often don\u2019t work well for images from other places with different socioeconomic characteristics. SEER delivers strong results for images from all around the globe \u2013 including non-U.S. and non-Europe regions with a wide range of income levels. SEER10B drastically improved performance on fairness benchmarks across gender, apparent skin tone, and age groups. Apart from its improved performance on fairness benchmarks, this model understands images from across the world well enough to localize them with unprecedented precision. We hope SEER will be an important building block as the AI community works to build systems that work well for everyone.<\/span><\/p>\n In August 2022, researchers in London and Paris open sourced the code for<\/span> Implicitron<\/span><\/a>, <\/span>a modular framework within our open source PyTorch3D library. Implictron uses neural implicit representation, a computer vision technique that can seamlessly combine real and virtual objects in augmented reality \u2014 without requiring large amounts of data to learn from and without being limited to just a few points of view.\u00a0<\/span><\/p>\n Implicitron learns a representation of a 3D object or scene using a sparse set of combined images of that object or scene from arbitrary viewpoints. Unlike traditional 3D representations such as meshes or point clouds, this newer approach represents objects as a continuous function, which allows for more accurate reconstruction of shapes with complex geometries as well as higher color reconstruction accuracy.<\/span><\/p>\n Our team in Tel Aviv is working closely on generative AI and has been at the forefront of some of Meta\u2019s most recent advancements. In July 2022, our Tel Aviv researchers and collaborators around the world <\/span>created a <\/span>generative AI research model called Make-A-Scene<\/span><\/a>. This multimodal generative AI method puts creative control in the hands of people who use it by allowing them to describe and illustrate their vision through both text descriptions and freeform sketches, resulting in surreal art, such as a hot dog flying through the sky and skyscrapers in the desert.<\/span><\/p>\n\n
\n
Groundbreaking Large Language Model Research<\/h2>\n
Self-supervised Computer Vision Research<\/h2>\n
Advancements in 3D Modeling<\/h2>\n
Generative AI for Images and Video<\/h2>\n