scispace - formally typeset
Search or ask a question
Institution

IBM

CompanyArmonk, New York, United States
About: IBM is a company organization based out in Armonk, New York, United States. It is known for research contribution in the topics: Layer (electronics) & Signal. The organization has 134567 authors who have published 253905 publications receiving 7458795 citations. The organization is also known as: International Business Machines Corporation & Big Blue.


Papers
More filters
Proceedings ArticleDOI
Steven J. Rennie1, Etienne Marcheret1, Youssef Mroueh1, Jarret Ross1, Vaibhava Goel1 
21 Jul 2017
TL;DR: This paper considers the problem of optimizing image captioning systems using reinforcement learning, and shows that by carefully optimizing systems using the test metrics of the MSCOCO task, significant gains in performance can be realized.
Abstract: Recently it has been shown that policy-gradient methods for reinforcement learning can be utilized to train deep end-to-end systems directly on non-differentiable metrics for the task at hand. In this paper we consider the problem of optimizing image captioning systems using reinforcement learning, and show that by carefully optimizing our systems using the test metrics of the MSCOCO task, significant gains in performance can be realized. Our systems are built using a new optimization approach that we call self-critical sequence training (SCST). SCST is a form of the popular REINFORCE algorithm that, rather than estimating a baseline to normalize the rewards and reduce variance, utilizes the output of its own test-time inference algorithm to normalize the rewards it experiences. Using this approach, estimating the reward signal (as actor-critic methods must do) and estimating normalization (as REINFORCE algorithms typically do) is avoided, while at the same time harmonizing the model with respect to its test-time inference procedure. Empirically we find that directly optimizing the CIDEr metric with SCST and greedy decoding at test-time is highly effective. Our results on the MSCOCO evaluation sever establish a new state-of-the-art on the task, improving the best result in terms of CIDEr from 104.9 to 114.7.

1,313 citations

Journal ArticleDOI
TL;DR: The Ball-Pivoting Algorithm is applied to datasets of millions of points representing actual scans of complex 3D objects and the quality of the results obtained compare favorably with existing techniques.
Abstract: The Ball-Pivoting Algorithm (BPA) computes a triangle mesh interpolating a given point cloud. Typically, the points are surface samples acquired with multiple range scans of an object. The principle of the BPA is very simple: Three points form a triangle if a ball of a user-specified radius p touches them without containing any other point. Starting with a seed triangle, the ball pivots around an edge (i.e., it revolves around the edge while keeping in contact with the edge's endpoints) until it touches another point, forming another triangle. The process continues until all reachable edges have been tried, and then starts from another seed triangle, until all points have been considered. The process can then be repeated with a ball of larger radius to handle uneven sampling densities. We applied the BPA to datasets of millions of points representing actual scans of complex 3D objects. The relatively small amount of memory required by the BPA, its time efficiency, and the quality of the results obtained compare favorably with existing techniques.

1,311 citations

Journal ArticleDOI
TL;DR: In this article, the authors present a review of the literature on high Ku alternative media, both for longitudinal and perpendicular recording, with data on sputtered and evaporated thin FePt films, with coercivities exceeding 10000 Oe.
Abstract: High K/sub u/, uniaxial magnetocrystalline anisotropy, materials are generally attractive for ultrahigh density magnetic recording applications as they allow smaller, thermally stable media grains. Prominent candidates are rare-earth transition metals (Co/sub 5/Sm,...), and tetragonal intermetallic compounds (L1/sub 0/ phases FePt, CoPtY,...), which have 20-40 times higher K/sub u/ than today's hexagonal Co-alloy based media. This allows for about 3 times smaller grain diameters, D, and a potential 10-fold areal density increase (/spl prop/1/D/sup 2/), well beyond the currently projected 40-100 Gbits/in/sup 2/ mark, Realization of such densities will depend on a large number of factors, not all related to solving media microstructure problems, In particular it is at present not known how to record into such media, which may require write fields in the order of 10-100 kOe. Despite this unsolved problem, there is considerable interest in high Ku alternative media, both for longitudinal and perpendicular recording. Activities in this area will be reviewed and data on sputtered and evaporated thin FePt films, with coercivities exceeding 10000 Oe will be presented.

1,310 citations

Book ChapterDOI
01 Jan 2014
TL;DR: The main ideas of agile software development processes were first introduced by the Agile Manifesto and second by presenting specific agile practices that enable agile teams to accomplish their development task on high quality.
Abstract: This chapter introduces the main ideas that form the basis for the agile approach Originally, the agile approach offers a professional approach for software development that encompasseshuman, organizational, and technological aspects of software development processes The main ideas of agile software development processes were first introduced by the Agile Manifesto and second by presenting specific agile practices that enable agile teams to accomplish their development task on high quality In the chapter, we present the Agile Manifesto as was published for software development and shows how it can be implemented for any projects

1,307 citations

Journal ArticleDOI
03 Nov 2005-Nature
TL;DR: An over 300-fold reduction of the group velocity on a silicon chip via an ultra-compact photonic integrated circuit using low-loss silicon photonic crystal waveguides that can support an optical mode with a submicrometre cross-section is experimentally demonstrated.
Abstract: It is known that light can be slowed down in dispersive materials near resonances. Dramatic reduction of the light group velocity-and even bringing light pulses to a complete halt-has been demonstrated recently in various atomic and solid state systems, where the material absorption is cancelled via quantum optical coherent effects. Exploitation of slow light phenomena has potential for applications ranging from all-optical storage to all-optical switching. Existing schemes, however, are restricted to the narrow frequency range of the material resonance, which limits the operation frequency, maximum data rate and storage capacity. Moreover, the implementation of external lasers, low pressures and/or low temperatures prevents miniaturization and hinders practical applications. Here we experimentally demonstrate an over 300-fold reduction of the group velocity on a silicon chip via an ultra-compact photonic integrated circuit using low-loss silicon photonic crystal waveguides that can support an optical mode with a submicrometre cross-section. In addition, we show fast (approximately 100 ns) and efficient (2 mW electric power) active control of the group velocity by localized heating of the photonic crystal waveguide with an integrated micro-heater.

1,307 citations


Authors

Showing all 134658 results

NameH-indexPapersCitations
Zhong Lin Wang2452529259003
Anil K. Jain1831016192151
Hyun-Chul Kim1764076183227
Rodney S. Ruoff164666194902
Tobin J. Marks1591621111604
Jean M. J. Fréchet15472690295
Albert-László Barabási152438200119
György Buzsáki15044696433
Stanislas Dehaene14945686539
Philip S. Yu1481914107374
James M. Tour14385991364
Thomas P. Russell141101280055
Naomi J. Halas14043582040
Steven G. Louie13777788794
Daphne Koller13536771073
Network Information
Related Institutions (5)
Carnegie Mellon University
104.3K papers, 5.9M citations

93% related

Georgia Institute of Technology
119K papers, 4.6M citations

92% related

Bell Labs
59.8K papers, 3.1M citations

90% related

Microsoft
86.9K papers, 4.1M citations

89% related

Massachusetts Institute of Technology
268K papers, 18.2M citations

88% related

Performance
Metrics
No. of papers from the Institution in previous years
YearPapers
202330
2022137
20213,163
20206,336
20196,427
20186,278