July 25, 2016

Single Image 3D Interpreter Network

European Conference on Computer Vision (ECCV)

Understanding 3D object structure from a single image is an important but difficult task in computer vision, mostly due to the lack of 3D object annotations in real images. Previous work tackles this problem by either solving an optimization task given 2D keypoint positions, or training on synthetic data with ground truth 3D information. In this work, we propose 3D INterpreter Network (3D-INN), an end-to-end framework which sequentially estimates 2D keypoint heatmaps and 3D object structure, trained on both real 2D-annotated images and synthetic 3D data.

Jiajun Wu, Tianfan Xue, Joseph J. Lim, Yuandong Tian, Joshua B. Tenenbaum, Antonio Torralba, William T. Freeman
July 23, 2016

HapticWave: Directional Surface Vibrations using Wave-Field Synthesis

SIGGRAPH 2016

HapticWave is a novel haptic technology that delivers directional haptic sensations generated on a flat surface to the user, without requiring him/her to wear a physical device.

Ravish Mehra, Chris Clock, David Perek, Elia Gatti, Riccardo DeSalvo, Sean Keller
July 19, 2016

Luminescent detector for free-space optical communication

Optica 3, 787-792 (2016)

We show that fluorescent materials can be used to increase the active area of a photodiode by orders of magnitude while maintaining its short response time and increasing its field of view.

Thibault Peyronel, Kevin Quirk, Tony Wang, Tobias Tiecke
June 27, 2016

Unsupervised Learning of Edges

CVPR

Data-driven approaches for edge detection have proven effective and achieve top results on modern benchmarks. However, all current data-driven edge detectors require manual supervision for training in the form of hand-labeled region segments or object boundaries.

Yin Li, Manohar Paluri, James M. Rehg, Piotr Dollar
June 25, 2016

Realtime Data Processing at Facebook

ACM SIGMOD

Realtime data processing powers many use cases at Facebook, including realtime reporting of the aggregated, anonymized voice of Facebook users, analytics for mobile applications, and insights for Facebook page administrators.

Guoqiang Jerry Chen, Janet Wiener, Shridhar Iyer, Anshul Jaiswal, Ran Lei, Nikhil Simha, Wei Wang, Kevin Wilfong, Tim Williamson, Serhat Yilmaz
June 20, 2016

Combining Two and Three-Way Embeddings Models for Link Prediction in Knowledge Bases

Journal of Artificial Intelligence Research, JAIR.org

This paper tackles the problem of endogenous link prediction for Knowledge Base completion.

Antoine Bordes, Nicolas Usunier, Alberto Garcia-Duran, Yves Grandvalet
June 19, 2016

Recurrent Orthogonal Networks and Long-Memory Tasks

International Conference on Machine Learning

This paper analyzes two synthetic datasets originally outlined in (Hochreiter and Schmidhuber, 1997) which are used to evaluate the ability of RNNs to store information over many time steps and explicitly construct RNN solutions to these problems.

Mikael Henaff, Arthur Szlam, Yann LeCun
June 18, 2016

Dynamo: Facebook’s Data Center-Wide Power Management System

ISCA 2016

In this paper, we describe Dynamo – a data center-wide power management system that monitors the entire power hierarchy and makes coordinated control decisions to safely and efficiently use provisioned data center power.

Qiang Wu, Qingyuan Deng, Lakshmi Ganesh, Chang-Hong Raymond Hsu, Yun Jin, Sanjeev Kumar, Bin Li, Justin Meza, Yee Jiun Song
June 18, 2016

Learning Physical Intuition of Block Towers by Example

International Conference on Machine Learning

Wooden blocks are a common toy for infants, allowing them to develop motor skills and gain intuition about the physical behavior of the world. In this paper, we explore the ability of deep feed-forward models to learn such intuitive physics.

Adam Lerer, Sam Gross, Rob Fergus
June 11, 2016

Treadmill: Attributing the Source of Tail Latency through Precise Load Testing and Statistical Inference

International Symposium on Computer Architecture

Managing tail latency of requests has become one of the primary challenges for large-scale Internet services. In this paper, we develop a methodology for statistically rigorous performance evaluation for server workloads.

Yunqi Zhang, David Meisner, Jason Mars, Lingjia Tang
June 8, 2016

Key-Value Memory Networks for Directly Reading Documents

EMNLP 2016

This paper introduces a new method, Key-Value Memory Networks, that makes reading documents more viable by utilizing different encodings in the addressing and output stages of the memory read operation.

Alexander Miller, Adam Fisch, Jesse Dodge, Amir-Hossein Karimi, Antoine Bordes, Jason Weston
May 22, 2016

The Social Ties of Immigrant Communities in the United States

WEBSCI

In this paper we study the composition of Facebook social networks of people who have moved from one country to another.

Amaç Herdağdelen, Bogdan State, Lada Adamic, Winter Mason
May 21, 2016

Continuous Deployment at Facebook and OANDA

ICSE 2016: 38th IEEE Conference on Software Engineering

This paper describes the continuous deployment practices at two very different firms: Facebook and OANDA, and shows that continuous deployment does not inhibit productivity or quality even in the face of substantial engineering team and code size growth.

Tony Savor, Mitchell Douglas, Michael Gentili, Laurie Williams, Kent Beck, Michael Stumm
May 17, 2016

Lost in propagation? Unfolding news cycles from the source

ICWSM '16

The news media play an important role in informing the public on current events. Yet it has been difficult to understand the comprehensiveness of news media coverage on an event and how the reactions that the coverage evokes may diverge, because this requires identifying the origin of an event and tracing the information all the way to individuals who consume the news.

Chenhao Tan, Adrien Friggeri, Lada Adamic
May 14, 2016

The shortest path is not always a straight line. Leveraging semi-metricity in graph analysis.

VLDB 2016

This paper leverages the concept of the metric backbone to improve the efficiency of large-scale graph analytics.

Dionysios Logothetis
May 7, 2016

Changes in Engagement Before and After Posting to Facebook

CHI 2016

We are interested in the ways in which posting content changes individuals’ engagement with Facebook. The study augments previous knowledge about uses and gratifications from posting on social network sites by focusing on short-term activity of contributors.

Nir Grinberg, Alex Dow, Lada Adamic, Mor Naaman
May 7, 2016

Legacy Contact: Designing and implementing stewardship at Facebook

CHI 2016

In this paper we describe Legacy Contact, a new feature that better supports people coping with the death of a loved one on Facebook while giving people more control over what happens to their Facebook account in the future.

Jed R. Brubaker, Vanessa Callison-Burch
May 7, 2016

Finexus: Tracking Fingertips Using Four Magnetometers

CHI 2016

We built a system to magnetically track the fingertips in order to bring hands into VR. It works by instrumenting electromagnets at the tips of the finger; the system is able to track the delicate movements in real time using only four magnetic sensors.

Ke-Yu Chen, Shwetak Patel, Sean Keller
May 2, 2016

Deep multi-scale video prediction beyond mean square error

ICLR 2016

The paper is about predicting future frames in video sequences given the previous frames.

Michael Mathieu, Camille Couprie, Yann LeCun
May 2, 2016

Predicting Distributions with Linearizing Belief Networks

ICLR 2016: International Conference on Learning Representation

This work introduces a new family of networks called linearizing belief nets or LBNs.

David Grangier