Research


Wo sind die Frauen in der IT? [DE]

Dr. Maria Börner - Adap GmbH
Podcast vom KI-Bundesverband

Diesmal sprechen wir mit Dr. Maria Börner. Sie ist Programm Managerin bei Adap GmbH und setzt sich neben ihren Job für mehr Frauen in Tech-Bereichen ein.


Federated Learning mit Flower [DE]

Daniel J. Beutel - Adap GmbH
Podcast vom KI-Bundesverband

Wir sprechen in der vierten Folge darüber was Federated Learning ist. Dabei stellt Daniel J. Beutel das Open Source Framework flower und erklärt die Vorteile von Federated Learning vor.


Flower Summit 2021 [EN]

Federated Learning & Flower Community
Recorded talks of the Flower Summit 2021

The Flower Summit includes future directions in federated learning research such as:

  • Future directions in Federated Learning
  • How to train federated speech recognition models using Flower and SpeechBrain
  • Can federated learning save the world? The carbon footprint of federated learning
  • Code tutorials showcasing both basic and advanced Flower workloads
  • Scaling your FL experiments and deployment tests to 1000s of clients
  • Recent academic research results built using the Flower framework


Federated Learning mit Dr. Maria Börner [DE]

Dr. Maria Börner - Adap GmbH
Recorded talk for "The Erium Podcast"

Die Nutzung von KI-Modellen nimmt weltweit zu. Oftmals sind die Daten auf mehreren Datenspeichern verteilt. Bei der herkömmlichen KI-Konfigurationen mit zentralem Training, wo die Daten in einer Cloud oder einem Server verarbeitet werden, gibt es jedoch mehrere Nachteile: Verletzung des Datenschutzes, hohe Latenzzeiten, komplexe Infrastruktur und hoher Energieverbrauch in Rechenzentren aufgrund der direkten Kühlung. Föderiertes Lernen bringt das KI-Training direkt zu den Datenspeichern und löst die genannten Probleme von vornherein... (Meetup Announcement)


Wake up and smell the Flower(s) [EN]

Daniel J. Beutel - Adap GmbH
Recorded talk for Meetup-AI Global

Present AI applications are running mostly in data centers or the cloud with a centralized training approach. However, this requires a complex big data infrastructure resulting in limited availability and high latency. Regulatory constraints are an additional challenge that needs to be solved. Federated learning addresses these problems by moving AI training to the data being saved on edge-devices... (Meetup Announcement)


Federated Learning For All With Flower [EN]

Daniel J. Beutel - Adap GmbH
Podcast recorded for The Python Podcast.__init__

Machine learning is a tool that has typically been performed on large volumes of data in one place. As more computing happens at the edge on mobile and low power devices, the learning is being federated which brings a new set of challenges. Daniel Beutel co-created the Flower framework to make federated learning more manageable. In this episode he shares his motivations for starting the project, how you can use it for your own work, and the unique challenges and benefits that this emerging model offers...


Fleet Intelligence with Federated Learning [EN]

Dr. Maria Börner - Adap GmbH
Zoom Recorded talk (Code: 8aiF6&Ky) for KI Community, 25 November 2020.

Present AI applications are running mostly in data centers or cloud with a centralized training approach. However, this requires a complex big data infrastructure resulting in limited availability and high latency. Regulatory constraints are an additional challenge that needs to be solved. Federated learning addresses these problems by moving AI training to the data being saved on decentralized edge-devices... (Meetup Announcement)


[NeurIPS 2020 Workshop] Can Federated Learning Save The Planet?

Xinchi Qiu, Titouan Parcollet, Daniel J. Beutel, Taner Topal, Akhil Mathur, Nicholas D. Lane
NeurIPS 2020 Workshop: Tackling Climate Change with Machine Learning
arXiv:2010.06537

Despite impressive results, deep learning-based technologies also raise severe privacy and environmental concerns induced by the training procedure often conducted in data centers. In response, alternatives to centralized training such as Federated Learning (FL) have emerged. Perhaps unexpectedly, FL in particular is starting to be deployed at a global scale by companies that must adhere to new legal demands and policies originating from governments and the civil society for privacy protection. However, the potential...


FLOW Seminar #23 [EN]

Dr. Nicholas Lane - University of Cambridge
Recorded talk for FLOW Seminar #23

Federated Learning One World Seminar, 4th November 2020, presents "Flower: A Friendly Federated Learning Framework, and a first look into the carbon footprint of federated methods".


Flower: A Friendly Federated Learning Research Framework

Daniel J. Beutel, Taner Topal, Akhil Mathur, Xinchi Qiu, Titouan Parcollet, Nicholas D. Lane
arXiv:2007.14390, Jul, 2020

Federated Learning (FL) has emerged as a promising technique for edge devices to collaboratively learn a shared prediction model, while keeping their training data on the device, thereby decoupling the ability to do machine learning from the need to store the data in the cloud. However, FL is difficult to implement and deploy in practice...


The Deep (Learning) Transformation of Mobile and Embedded Computing [EN]

Dr. Nicholas Lane - University of Cambridge
Recorded talk for tinyML Talks local Webcast, 8 September 2020.

Mobile and embedded devices increasingly rely on deep neural networks to understand the world -- a formerly impossible feat that would have overwhelmed their system resources just a few years ago. The age of on-device artificial intelligence is upon us; but incredibly, these dramatic changes are just the beginning... (Meetup Announcement)


Flower Open Source Project

Flower is our open source federated learning framework. It offers a unified approach to federated learning. Federate any workload, any ML framework, and any programming language.