机器之心海外团队作者

Synced Global AI Weekly | 2018.9.8—9.14

Synced's Favourite Tech News of This Week

Tencent Open-Sources Its Massive Multi-Labeled Image Dataset

Tencent AI Lab has announced that it will open source its multi-label image dataset ML-Images and deep residual network ResNet-101 by the end of September. ML-Images contains 18 million images and more than 11,000 common object categories; while ResNet-101 has reached the highest precision level in the industry.

(Synced)


AMC: AutoML for Model Compression and Acceleration on Mobile Devices 

Researchers from MIT, Google, and Xian Jiaotong University recently published a paper proposing AutoML for Model Compression (AMC), which leverages reinforcement learning to shorten model compression processing time and improve results.

(MIT & Google)


Is It Real or Is It Gibson? A New Virtual Environment for AI Training

Researchers from Stanford University and University of California Berkeley have introduced Gibson Environment, a real-world-based virtual environment for training and testing active perception agents.

(Synced)


ECCV 2018 Announces Best Papers

Taking top honours in the Best Paper category is Implicit 3D Orientation Learning for 6D Object Detection from RGB Images from researchers at the German Aerospace Center andTechnical University of Munich.  

(Synced)


Technology

Preserving Outputs Precisely while Adaptively Rescaling Targets

"PopArt: a single agent that can play 57 diverse Atari video games, with above-human median performance across the set"

(DeepMind)


Finding And Fixing Software Bugs Automatically with SapFix and Sapienz

Debugging code is drudgery. But SapFix, a new AI hybrid tool created by Facebook engineers, can significantly reduce the amount of time engineers spend on debugging, while also speeding up the process of rolling out new software. SapFix can automatically generate fixes for specific bugs, and then propose them to engineers for approval and deployment to production.

(Facebook)


The What-If Tool: Code-Free Probing of Machine Learning Models

"Today, we are launching the What-If Tool, a new feature of the open-source TensorBoard web application, which let users analyze and better understand an ML model without writing code. We look forward to people using, and contributing to, the What-If Tool."

(Google AI)


You May Also Like

AI Chip Duel: Apple A12 Bionic VS Huawei Kirin 980

Apple has unveiled the latest iteration of its smartphone chip: the A12 Bionic SoC (system-on-a-chip). The company made the announcement yesterday at its annual product showcase event in Cupertino, California, hailing the A12 as the industry’s first ever 7nm chip (the smallest current transistor scale). It will be embedded in Apple's new XR, XS, and XS Max iPhones.

(Synced


Jeff Dean's 1990 Senior Thesis Is Better Than Yours

Google AI lead Jeff Dean recently posted a link to his 1990 senior thesis on Twitter, which set off a wave of nostalgia for the early days of machine learning in the AI community. The thesis may be almost 30 years old and only eight pages long, but the paper does a remarkable job of explaining the methods behind neural network training and the modern development of AI.

(Synced)


Global AI Events

17–18 Sep
ICHRI
Rome, Italy.
17–18 Sep
AI Innovation Summit.
San Francisco, USA.
17–19 Sep
AIPR 2018
 Lodz, Poland.
18–20 Sept
IJCCI
Seville, Spain.
18–20 Sep
AI Summit
San Francisco, USA.
19–21 Sep
CHIRA
Seville, Spain
NewsLetter
相关数据
机器学习技术

机器学习是人工智能的一个分支,是一门多领域交叉学科,涉及概率论、统计学、逼近论、凸分析、计算复杂性理论等多门学科。机器学习理论主要是设计和分析一些让计算机可以自动“学习”的算法。因为学习算法中涉及了大量的统计学理论,机器学习与推断统计学联系尤为密切,也被称为统计学习理论。算法设计方面,机器学习理论关注可以实现的,行之有效的学习算法。

感知技术

知觉或感知是外界刺激作用于感官时,脑对外界的整体的看法和理解,为我们对外界的感官信息进行组织和解释。在认知科学中,也可看作一组程序,包括获取信息、理解信息、筛选信息、组织信息。与感觉不同,知觉反映的是由对象的各样属性及关系构成的整体。

迭代 技术

模型的权重在训练期间的一次更新。迭代包含计算参数在单个批量数据上的梯度损失。

TensorBoard技术

一个信息中心,用于显示在执行一个或多个 TensorFlow 程序期间保存的摘要信息。

张量技术

张量是一个可用来表示在一些矢量、标量和其他张量之间的线性关系的多线性函数,这些线性关系的基本例子有内积、外积、线性映射以及笛卡儿积。其坐标在 维空间内,有 个分量的一种量,其中每个分量都是坐标的函数,而在坐标变换时,这些分量也依照某些规则作线性变换。称为该张量的秩或阶(与矩阵的秩和阶均无关系)。 在数学里,张量是一种几何实体,或者说广义上的“数量”。张量概念包括标量、矢量和线性算子。张量可以用坐标系统来表达,记作标量的数组,但它是定义为“不依赖于参照系的选择的”。张量在物理和工程学中很重要。例如在扩散张量成像中,表达器官对于水的在各个方向的微分透性的张量可以用来产生大脑的扫描图。工程上最重要的例子可能就是应力张量和应变张量了,它们都是二阶张量,对于一般线性材料他们之间的关系由一个四阶弹性张量来决定。

推荐文章
暂无评论
暂无评论~