All Research Areas
Research Areas
Year Published

387 Results

April 15, 2018

Learning Filterbanks from Raw Speech for Phone Recognition

International Conference on Acoustics, Speech and Signal Processing (ICASSP)

We train a bank of complex filters that operates on the raw waveform and is fed into a convolutional neural network for end-to-end phone recognition.

By: Neil Zeghidour, Nicolas Usunier, Iasonas Kokkinos, Thomas Schatz, Gabriel Synnaeve, Emmanuel Dupoux
April 15, 2018

Towards End-to-End Spoken Language Understanding

International Conference on Acoustics, Speech and Signal Processing (ICASSP 2018)

Spoken language understanding system is traditionally designed
as a pipeline of a number of components.

By: Dmitriy Serdyuk, Yongqiang Wang, Christian Fuegen, Anuj Kumar, Baiyang Liu, Yoshua Bengio
March 28, 2018

SmoothOperator: Reducing Power Fragmentation and Improving Power Utilization in Large-scale Datacenters


With the ever growing popularity of cloud computing and web services, Internet companies are in need of increased computing capacity to serve the demand. However, power has become a major limiting factor prohibiting the growth in industry: it is often the case that no more servers can be added to datacenters without surpassing the capacity of the existing power infrastructure.

By: Chang-Hong Hsu, Qingyuan Deng, Jason Mars, Lingjia Tang
March 25, 2018

Toward Improved Surgical Training: Delivering Smoothness Feedback using Haptic Cues

IEEE Haptics Symposium (HAPTICS)

In this paper, we explore the effectiveness of providing real-time feedback of movement smoothness, a characteristic associated with skilled and coordinated movement, via a vibrotactile cue. Subjects performed a mirror-tracing task that requires coordination and dexterity similar in nature to that required in endovascular surgery.

By: William H. Jantscher, Shivam Pandey, Priyanshu Agarwal, Sadie H. Richardson, Bowie R. Lin, Michael D. Byrne, Marcia K. O’Malley
March 25, 2018

A Social Haptic Device to Create Continuous Lateral Motion Using Sequential Normal Indentation

IEEE Haptics Symposium (HAPTICS)

This paper presents a device for creating a continuous lateral motion on the arm to mimic a subset of the gestures used in social touch.

By: Heather Culbertson, Cara M. Nunez, Ali Israr, Frances Lau, Freddy Abnousi, Allison M. Okamura
February 2, 2018

StarSpace: Embed All The Things!

Conference on Artificial Intelligence (AAAI)

We present StarSpace, a general-purpose neural embedding model that can solve a wide variety of problems: labeling tasks such as text classification, ranking tasks such as information retrieval/web search, collaborative filtering-based or content-based recommendation, embedding of multi-relational graphs, and learning word, sentence or document level embeddings.

By: Ledell Wu, Adam Fisch, Sumit Chopra, Keith Adams, Antoine Bordes, Jason Weston
February 2, 2018

Efficient Large-Scale Multi-Modal Classification

Conference on Artificial Intelligence (AAAI)

We investigate various methods for performing multi-modal fusion and analyze their trade-offs in terms of classification accuracy and computational efficiency.

By: Douwe Kiela, Edouard Grave, Armand Joulin, Tomas Mikolov
January 27, 2018

An Inverse-Kinematic Approach to Dual-Stage Servo Control for an Optical Pointing System

SPIE Photonics West

In free-space optical communication systems, narrow beam diameters necessitate precision pointing over a relatively wide field of regard. Control systems with such large dynamic range requirements sometimes employ multi-stage architectures, where a “coarse” mechanism provides low-bandwidth control over a large range of travel, and a “fine” mechanism provides high-bandwidth disturbance rejection over a limited range of travel. In such systems, the two stages must be coordinated in their dynamic response, so as to avoid undesirable coupling and even interference with each other. This topic has been studied in various literature – especially in the field of hard disk drives, but also for optical pointing systems. However, most of this literature considers systems where the output of the two stages combines by simple linear sum. Dual-stage control becomes even more challenging when the two stages combine via nonlinear kinematic relationships, as they would in an optical pointing system employing multiple gimbaled steering mirrors. This paper presents an approach for handling these nonlinear kinematic effects, and demonstrates the viability of this approach via simulation.

By: Eric D. Miller
January 14, 2018

Experimental Demonstration of Digital Pre-Distortion for Millimeter Wave Power Amplifiers with GHz Bandwidth

IEEE Radio Wireless Week (RWW)

This paper presents an experimental demonstration of digital pre-distortion (DPD) on E-band power amplifiers (PA) with GHz channel bandwidth.

By: Qi Tang, Hongyu Zhou, Abhishek Tiwari, Joseph Stewart, Qi Qu, Dawei Zhang, Hamid Hemmati
December 12, 2017

Supporting Diverse Dynamic Intent-based Policies using Janus

International Conference on emerging Networking EXperiments and Technologies (CoNEXT)

In this paper we propose Janus, a system which makes two major contributions to network policy abstractions. First, we extend the prior policy graph abstraction model to represent complex QoS and dynamic stateful/temporal policies. Second, we convert the policy configuration problem into an optimization problem with the goal of maximizing the number of satisfied and configured policies, and minimizing the number of path changes under dynamic environments.

By: Anubhavnidhi Abhashkumar, Joon-Myung Kang, Sujata Banerjee, Aditya Akella, Ying Zhang, Wenfei Wu