Research Area
Year Published

932 Results

January 1, 2020

Designing Safe Spaces for Virtual Reality

Ethics in Design and Communication

Virtual Reality (VR) designers accept the ethical responsibilities of removing a user’s entire world and superseding it with a fabricated reality. These unique immersive design challenges are intensified when virtual experiences become public and socially-driven. As female VR designers in 2018, we see an opportunity to fold the language of consent into the design practice of virtual reality—as a means to design safe, accessible, virtual spaces.

Publication will be made available in 2020.

By: Michelle Cortese, Andrea Zeller

December 14, 2019

From Senones to Chenones: Tied Context-Dependent Graphemes for Hybrid Speech Recognition

IEEE Automatic Speech Recognition and Understanding Workshop

There is an implicit assumption that traditional hybrid approaches for automatic speech recognition (ASR) cannot directly model graphemes and need to rely on phonetic lexicons to get competitive performance, especially on English which has poor grapheme-phoneme correspondence. In this work, we show for the first time that, on English, hybrid ASR systems can in fact model graphemes effectively by leveraging tied context-dependent graphemes, i.e., chenones.

By: Duc Le, Xiaohui Zhang, Weiyi Zhang, Christian Fuegen, Geoffrey Zweig, Michael L. Seltzer

November 10, 2019

Adversarial Bandits with Knapsacks

Symposium on Foundations of Computer Science (FOCS)

We consider Bandits with Knapsacks (henceforth, BwK), a general model for multi-armed bandits under supply/budget constraints. In particular, a bandit algorithm needs to solve a well-known knapsack problem: find an optimal packing of items into a limited-size knapsack. The BwK problem is a common generalization of numerous motivating examples, which range from dynamic pricing to repeated auctions to dynamic ad allocation to network routing and scheduling.

By: Nicole Immorlica, Karthik Abinav Sankararaman, Robert Schapire, Aleksandrs Slivkins

November 10, 2019

EASSE: Easier Automatic Sentence Simplification Evaluation

Conference on Empirical Methods in Natural Language Processing (EMNLP)

We introduce EASSE, a Python package aiming to facilitate and standardize automatic evaluation and comparison of Sentence Simplification (SS) systems. EASSE provides a single access point to a broad range of evaluation resources: standard automatic metrics for assessing SS outputs (e.g. SARI), word-level accuracy scores for certain simplification transformations, reference-independent quality estimation features (e.g. compression ratio), and standard test data for SS evaluation (e.g. TurkCorpus).

By: Fernando Alva-Manchego, Louis Martin, Carolina Scarton, Lucia Specia

November 10, 2019

An Integrated 6DoF Video Camera and System Design

SIGGRAPH Asia

Designing a fully integrated 360◦ video camera supporting 6DoF head motion parallax requires overcoming many technical hurdles, including camera placement, optical design, sensor resolution, system calibration, real-time video capture, depth reconstruction, and real-time novel view synthesis. While there is a large body of work describing various system components, such as multi-view depth estimation, our paper is the first to describe a complete, reproducible system that considers the challenges arising when designing, building, and deploying a full end-to-end 6DoF video camera and playback environment.

By: Albert Parra Pozo, Michael Toksvig, Terry Filiba Schrager, Joyce Hsu, Uday Mathur, Alexander Sorkine-Hornung, Richard Szeliski, Brian Cabral

November 9, 2019

Harassment in Social Virtual Reality: Challenges for Platform Governance

Conference on Computer-Supported Cooperative Work and Social Computing (CSCW)

In immersive virtual reality (VR) environments, experiences of harassment can be exacerbated by features such as synchronous voice chat, heightened feelings of presence and embodiment, and avatar movements that can feel like violations of personal space (such as simulated touching or grabbing). Simultaneously, efforts to govern these developing spaces are made more complex by the distributed landscape of virtual reality applications and the dynamic nature of local community norms. To better understand this nascent social and psychological environment, we interviewed VR users (n=25) about their experiences with harassment, abuse, and discomfort in social VR.

By: Lindsay Blackwell, Nicole Ellison, Natasha Elliott-Deflo, Raz Schwartz

November 7, 2019

Feature Selection for Facebook Feed Ranking System via a Group-Sparsity-Regularized Training Algorithm

Conference on Information and Knowledge Management (CIKM)

In modern production platforms, large scale online learning models are applied to data of very high dimension. To save computational resource, it is important to have an efficient algorithm to select the most significant features from an enormous feature pool. In this paper, we propose a novel neural-network-suitable feature selection algorithm, which selects important features from the input layer during training.

By: Xiuyan Ni, Yang Yu, Peng Wu, Youlin Li, Shaoliang Nie, Qichao Que, Chao Chen

November 7, 2019

Cloze-driven Pretraining of Self-attention Networks

Conference on Empirical Methods in Natural Language Processing (EMNLP)

We present a new approach for pretraining a bi-directional transformer model that provides significant performance gains across a variety of language understanding problems.

By: Alexei Baevski, Sergey Edunov, Yinhan Liu, Luke Zettlemoyer, Michael Auli

November 5, 2019

Revisiting the Evaluation of Theory of Mind through Question Answering

Conference on Empirical Methods in Natural Language Processing (EMNLP)

Theory of mind, i.e., the ability to reason about intents and beliefs of agents is an important task in artificial intelligence and central to resolving ambiguous references in natural language dialogue. In this work, we revisit the evaluation of theory of mind through question answering.

By: Matthew Le, Y-Lan Boureau, Maximilian Nickel

November 5, 2019

A Discrete Hard EM Approach for Weakly Supervised Question Answering

Conference on Empirical Methods in Natural Language Processing (EMNLP)

Many question answering (QA) tasks only provide weak supervision for how the answer should be computed. For example, TRIVIAQA answers are entities that can be mentioned multiple times in supporting documents, while DROP answers can be computed by deriving many different equations from numbers in the reference text. In this paper, we show it is possible to convert such tasks into discrete latent variable learning problems with a precomputed, task-specific set of possible solutions (e.g. different mentions or equations) that contains one correct option.

By: Sewon Min, Danqi Chen, Hannaneh Hajishirzi, Luke Zettlemoyer