aihub.org
Last month in tweets – October 2019
We have collected some of the month’s most interesting tweets about AI.
Efforts to push back harmful AI in one year
"Those of us concerned with the implications of AI need to seek out and amplify the people already doing the work and learn the histories of those who’ve led the way." – @mer__edith outlines how supporters & partners can push back against harmful AI at #AINow2019 @VaroonMathur pic.twitter.com/BLWpbFrex4
— AI Now Institute (@AINowInstitute) October 2, 2019
Privacy-preserving learning system for medical imaging analysis
.@KingsCollegeLon and NVIDIA #Research developed the first privacy-preserving federated learning system for medical imaging analysis. Learn more about this breakthrough in #healthcare #AI.
— NVIDIA AI (@NvidiaAI) October 14, 2019
An open-source platform to promote research and development on real-world robotics hardware
Check out ROBEL, an open-source platform for cost-effective robots and curated benchmarks designed to facilitate research and development on real-world hardware. Learn more below ↓ https://t.co/m5WWTfCDAp
— Google AI (@GoogleAI) October 9, 2019
Paraphrase Adversaries from Word Scrambling (PAWS) English dataset and multilingual PAWS-X for better multilingual NLP models
Today, we are excited to release the English language PAWS (Paraphrase Adversaries from Word Scrambling) and multilingual PAWS-X datasets, to help the community develop multilingual models that better exploit structure, context and pairwise comparisons. https://t.co/kKVKXjWbMX
— Google AI (@GoogleAI) October 2, 2019
Research work on an end-to-end real-time multilingual speech recognition system
High-quality #automaticspeechrecognition systems require large amounts of data—yet many languages have little data available. Check out new research into an end-to-end system trained as a single model allowing for real-time multilingual speech recognition. https://t.co/OAwwIDCk78
— Google AI (@GoogleAI) September 30, 2019
A software development kit for building AI applications that merge between vision, speech, and other sensors in order to make conversations more natural
Announced today at #MWC19 – NVIDIA Jarvis, an SDK for building and deploying #AI applications that fuse vision, speech and other sensors to enable conversation-based experiences that are natural, engaging and accurate. Apply for early access today: https://t.co/41UYIqfLP0
— NVIDIA AI Developer (@NVIDIAAIDev) October 21, 2019
A unified model for vision-language generation tasks
People learn directly and indirectly, accumulating commonsense knowledge to draw on. For machines, it’s more challenging. To better mimic human scene/language learning, researchers have developed VLP, a unified model leveraging vision-language pre-training https://t.co/SGPVsQKTN6
— Microsoft Research (@MSFTResearch) October 8, 2019
Robots with dexterous hands solving Rubik’s Cube 60% of the time (20% in case of a particularly hard scramble) using reinforcement learning and Kociemba’s algorithm for choosing the steps
We’re all used to robots that fail when their environment changes unpredictably. Our robotic system is adaptable enough to handle unexpected situations not seen during training, such as being prodded by a stuffed giraffe: pic.twitter.com/wBoh1nt9Kv
— OpenAI (@OpenAI) October 15, 2019
Addressing bias in AI
Addressing the lack of diversity and inclusion in AI is a community effort. Big thanks to @gwenmoran @FortuneMagazine for including AI4ALL as a key org making an impact on this issue alongside @Accenture, @microsoft, @IEEEorg, @PartnershipAI + more. https://t.co/8zOnH9WO9l
— AI4ALL (@ai4allorg) October 29, 2019
Restoring ancient inscriptions by recovering missing characters from damaged text input using deep learning
Could machine learning aid historians in the textual preservation and restoration of inscriptions from ancient Greece? Pythia is the first ancient text restoration model that recovers missing characters from a damaged text input using deep neural networks: https://t.co/PSBgQR349v pic.twitter.com/WqKBzlBUV2
— DeepMind (@DeepMindAI) October 17, 2019
AIhub is supported by:
