AI AI Weekly

Alibaba, Baidu and Xiaomi Open-Source Their SOTA Research

Synced Global AI Weekly July 7th

Subscribe to Synced Global AI Weekly


Alibaba DAMO Academy Open-Sources a New Generation of Human-Machine Dialogue Model: ESIM.
The Enhanced Sequential Inference Model (ESIM) has won double championships in the International Top Dialogue System Evaluation Competition, which increased the world record for human-machine dialogue accuracy to 94.1%. ESIM model-based applications such as subway voice ticketing machine have been launched in China.
(Alibaba DAMO Academy)/ (GitHub)


Baidu Open-Sources Apollo 5.0
Apollo 5.0 is an effort to support volume production for Geo-Fenced Autonomous Driving. The car now has 360-degree visibility, along with upgraded perception deep learning model to handle the changing conditions of complex road scenarios, making the car more secure and aware.
(GitHub)


XiaoMi AI Lab Open-Sources FairNAS: Rethinking Evaluation Fairness of Weight Sharing Neural Architecture Search
Researchers propose a novel idea called Fair Neural Architecture Search (FairNAS), in which a strict fairness constraint is enforced for fair inheritance and training. In this way, their supernet exhibits nice convergence and very high training accuracy.
(Xiaomi AI Lab) / (GitHub)

Technology

Neural Decipherment via Minimum-Cost Flow: from Ugaritic to Linear B
Researchers propose a novel neural approach for automatic decipherment of lost languages. To compensate for the lack of strong supervision signal, their model design is informed by patterns in language change documented in historical linguistics. The model utilizes an expressive sequence-to-sequence model to capture character-level correspondences between cognates.
(CSAIL MIT & Google Brain)


DLRM: An Advanced, Open Source Deep Learning Recommendation Model
Facebook AI is open-sourcing a state-of-the-art deep learning recommendation model that was implemented using Facebook’s open source PyTorch and Caffe2 platforms. DLRM advances on other models by combining principles from both collaborative filtering and predictive analytics-based approaches.
(Facebook AI)


Benchmarking Model-Based Reinforcement Learning
To facilitate research in Model-based reinforcement learning (MBRL), in this paper researchers gather a wide collection of MBRL algorithms and propose over 18 benchmarking environments specially designed for MBRL. They benchmark these algorithms with unified problem settings, including noisy environments.
(University of Toronto, Vector Institue and UC Berkeley)

You May Also Like

Baidu PaddlePaddle DL Framework & Huawei Kirin SoC: A Formidable Partnership
Baidu’s homegrown deep learning framework PaddlePaddle will empower Huawei’s Kirin smartphone chips, the company announced at the Baidu Create 2019 AI Developer Conference in Beijing.
(Synced)


Natural Beauty Meets Artificial Intelligence
The global cosmetic products market is expected to climb to US$863 billion in 2024 — an impressive CAGR of above 7 percent. AI-driven tech has shown it can provide highly personalized services and open new opportunities for beauty brands to engage with the public.
(Synced)

Global AI Events

August 19-23: Knowledge Discovery and Data Mining (KDD2019) in London, United Kingdom

September 10-12: The AI Summit (Part of TechXLR8) in Singapore

September 24-28: Microsoft Ignite in Orlando, United States

October 27-November 3: International Conference on Computer Vision (ICCV) in Seoul, South Korea

Global AI Opportunities

Research Scientist, Google Brain Toronto

OpenAI Seeking Software Engineers and Deep Learning Researchers

DeepMind is Recruiting

DeepMind Scholarship: Access to Science

Postdoctoral Researcher (AI) – Self-Supervised Learning

LANDING AI is recruiting


Stay tight with AI! 
Subscribe to Synced Global AI Weekly

1 comment on “Alibaba, Baidu and Xiaomi Open-Source Their SOTA Research

  1. Pingback: #sharing and #collaboration in #china and they call it “open source… | Dr. Roy Schestowitz (罗伊)

Leave a Reply

Your email address will not be published.

%d bloggers like this: