Evolving AI Lab

[Evolving AI Lab @EvolvingAI]

I am excited to give a talk at the ICML Workshop on Multi-Task and Lifelong Reinforcement Learning. Thanks to the organizers for the invitation: Chelsea Finn, Amy Zhang, Abhishek Gupta, Andrei Rusu, Khimya... https://t.co/5ufwf3mtI4

[Evolving AI Lab @EvolvingAI]

Our open-access Nature Machine Intelligence review is out: Designing neural networks through neuroevolution. https://t.co/BfXzhvqEht with... https://t.co/BfXzhvqEht

[Evolving AI Lab @EvolvingAI]

Creating a Zoo of Atari-Playing Agents to Catalyze the Understanding of Deep Reinforcement Learning. Great work led by Joel Lehman, with many excellent collaborators from Uber AI Labs, OpenAI, & Google Brain.

Blog:... https://t.co/mdIpSGQvT9

[Evolving AI Lab @EvolvingAI]

Introducing POET: it generates its own increasingly complex, diverse training environments and solves them. It automatically creates a learning curricula, its own training data, and creates, collects, and harnesses stepping... https://t.co/CgRZdgwOKA

[Evolving AI Lab @EvolvingAI]

https://t.co/8O88rfSdah https://t.co/8O88rfSdah

[Evolving AI Lab @EvolvingAI]

More surprisingly good results on deep learning for camera trap images! Proud to be a part of this work, led by the excellent Michael Tabak. "Machine learning to classify animal species in camera trap images" in... https://t.co/meLlmVRPGQ

[Evolving AI Lab @EvolvingAI]

A wide-ranging interview on AI, research at @UberAILabs, my background, and my thoughts on the future directions of AI conducted at the ReWork Conference.... https://t.co/DoyQc4sWKc

[Evolving AI Lab @EvolvingAI]

Our paper was accepted to NIPS! "Improving Exploration in Evolution Strategies for Deep Reinforcement Learning via a Population of Novelty-Seeking Agents". Congrats to leads Vashisht Madhavan, Edoardo Conti, and the... https://t.co/XgoCUb58S4

[Evolving AI Lab @EvolvingAI]

Learning general, multi-modal maze solvers. They learn the age-old general maze solution: follow a wall until you see the goal, then go to the goal. Excellent work by Joost Huizinga via the Combinatorial MultiObjective Evolutionary Algorithm... https://t.co/Mg53qhIKdh

[Evolving AI Lab @EvolvingAI]

Generating multi-modal robot behavior (neural networks that perform many different tasks well) via the new Combinatorial Multi-Objective Evolutionary Algorithm (CMOEA). I am very proud of this work and think some... https://t.co/mIdoWOFJOg

[Evolving AI Lab @EvolvingAI]

New paper & dataset (3.8M labeled images!): Deep learning works well at recognizing animals in camera trap images. Building on our PNAS paper w/ African animals, here w/ North American animals we get 98% accuracy! Led by... https://t.co/nHcmXsqZRE

[Evolving AI Lab @EvolvingAI]

A BBC video about our recent PNAS paper on using deep neural networks to automatically identify, count, and describe the behaviors of animals in motion-sensor camera images.

Congrats to current Evolving AI Lab... https://t.co/RqC06Yrxg7

[Evolving AI Lab @EvolvingAI]

A nice summary of our PNAS paper on automatically extracting data from camera trap images with deep neural networks.

https://t.co/CYBREoQ2ga https://t.co/CYBREoQ2ga

[Sen. Sally Eaves @sallyeaves]

Brilliant! How #AI is transforming fields that support #conservation from ecology to animal behaviour. This exemplar project uses AI to identify + count wildlife with 96.6% accuracy https://t.co/03ttxknhtW #BigData #innovation #TechForGood #nature #MachineLearning @itknowingness

[Evolving AI Lab @EvolvingAI]

https://t.co/AUpKJvv1rp https://t.co/AUpKJvv1rp

[Evolving AI Lab @EvolvingAI]

We are thrilled to announce our deep learning + motion-sensor camera work has been published in PNAS! “Automatically identifying, counting, and describing wild animals in camera-trap images with deep learning” We think... https://t.co/8Kb7uWqn2h

[Evolving AI Lab @EvolvingAI]

Fun to see our paper mentioned in the New Yorker. "The Surprising Creativity of Digital Evolution" informs their discussion titled "How Frightened Should We Be of A.I.?" with Joel Lehman Dule Misevic and a long group of amazing coauthors... https://t.co/bgxeTcnheL

[Evolving AI Lab @EvolvingAI]

Video of a talk by Jeff Clune at UC Merced recently. https://t.co/7bv0dczNzJ Title: A talk in two parts: (1) AI Neuroscience: How much do deep neural networks understand about the images they classify? (2) Robots... https://t.co/7bv0dczNzJ

[Evolving AI Lab @EvolvingAI]

Accelerating Deep Neuroevolution: Train Atari in Hours on a Single Personal Computer! What took ~1 hour on 720 CPUs now takes only ~4 hours on a *single* modern desktop. Code is open source. Awesome work by Felipe Such! With... https://t.co/zE2vvK7sVt

[Evolving AI Lab @EvolvingAI]

A huge congratulations to former EvolvingAI Lab PhD student Anh Nguyen, who just won the University of Wyoming 2018 Outstanding Dissertation of the Year award! It is a very well-deserved honor! Anh is now an assistant... https://t.co/NjNONW0m1X