Entity

Time filter

Source Type

Berkeley, CA, United States

Raghavan B.,ICSI | Ma J.,University of California at Berkeley
Proceedings of the 2nd ACM SIGCOMM Workshop on Green Networking, GreenNets'11 | Year: 2011

We explore responses to a scenario in which the severity of a permanent energy crisis fundamentally limits our ability to maintain the current-day Internet architecture. In this paper, we review why this scenario - whose vague outline is known to many but whose consequences are generally understood only by the scientists who study it - is likely, and articulate the specific impacts that it would have on network infrastructure and networking research. In light of this, we propose a concrete research agenda to address the networking needs of an energy-deprived society. © 2011 ACM.


Schuller B.,TU Munich | Steidl S.,ICSI | Steidl S.,Friedrich - Alexander - University, Erlangen - Nuremberg | Batliner A.,Friedrich - Alexander - University, Erlangen - Nuremberg | And 2 more authors.
Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH | Year: 2011

While the first open comparative challenges in the field of paralinguistics targeted more 'conventional' phenomena such as emotion, age, and gender, there still exists a multiplicity of not yet covered, but highly relevant speaker states and traits. The INTERSPEECH 2011 Speaker State Challenge thus addresses two new sub-challenges to overcome the usually low compatibility of results: In the Intoxication Sub-Challenge, al-coholisation of speakers has to be determined in two classes; in the Sleepiness Sub-Challenge, another two-class classification task has to be solved. This paper introduces the conditions, the Challenge corpora "Alcohol Language Corpus" and "Sleepy Language Corpus", and a standard feature set that may be used. Further, baseline results are given. Copyright © 2011 ISCA.


Allman M.,ICSI
Computer Communication Review | Year: 2015

Understanding, measuring, and debugging IP networks, particularly across administrative domains, is challenging. One particularly daunting aspect of the challenge is the presence of transparent middleboxes which are now common in today's Internet. In-path middleboxes that modify packet headers are typically transparent to a TCP, yet can impact end-to-end performance or cause blackholes. We develop TCP HICCUPS to reveal packet header manipulation to both endpoints of a TCP connection. HICCUPS permits endpoints to cooperate with currently opaque middleboxes without prior knowledge of their behavior. For example, with visibility into end-to-end behavior, a TCP can selectively enable or disable performance enhancing options. This cooperation enables protocol innovation by allowing new IP or TCP functionality (e.g., ECN, SACK, Multipath TCP, Tcpcrypt) to be deployed without fear of such functionality being misconstrued, modified, or blocked along a path. HICCUPS is incrementally deployable and introduces no new options. We implement and deploy TCP HICCUPS across thousands of disparate Internet paths, highlighting the breadth and scope of subtle and hard to detect middlebox behaviors encountered. We then show how path diagnostic capabilities provided by HICCUPS can benefit applications and the network. Copyright 2014 ACM.


Sundaresan S.,ICSI | Feamster N.,Princeton University | Teixeira R.,French Institute for Research in Computer Science and Automation
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | Year: 2015

This paper studies how home wireless performance characteristics affect the performance of user traffic in real homes. Previous studies have focused either on wireless metrics exclusively, without connection to the performance of user traffic; or on the performance of the home network at higher layers. In contrast, we deploy a passive measurement tool on commodity access points to correlate wireless performance metrics with TCP performance of user traffic. We implement our measurement tool, deploy it on commodity routers in 66 homes for one month, and study the relationship between wireless metrics and TCP performance of user traffic. We find that, most of the time, TCP flows from devices in the home achieve only a small fraction of available access link throughput; as the throughput of user traffic approaches the access link throughput, the characteristics of the home wireless network more directly affect performance.We also find that the 5 GHz band offers users better performance better than the 2.4GHz band, and although the performance of devices varies within the same home, many homes do not have multiple devices sending high traffic volumes, implying that certain types of wireless contention may be uncommon in practice. © Springer International Publishing Switzerland 2015.


Ghahremani P.,Johns Hopkins University | Babaali B.,University of Tehran | Povey D.,Johns Hopkins University | Riedhammer K.,ICSI | And 2 more authors.
ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings | Year: 2014

In this paper we present an algorithm that produces pitch and probability-of-voicing estimates for use as features in automatic speech recognition systems. These features give large performance improvements on tonal languages for ASR systems, and even substantial improvements for non-tonal languages. Our method, which we are calling the Kaldi pitch tracker (because we are adding it to the Kaldi ASR toolkit), is a highly modified version of the getf0 (RAPT) algorithm. Unlike the original getf0 we do not make a hard decision whether any given frame is voiced or unvoiced; instead, we assign a pitch even to unvoiced frames while constraining the pitch trajectory to be continuous. Our algorithm also produces a quantity that can be used as a probability of voicing measure; it is based on the normalized autocorrelation measure that our pitch extractor uses. We present results on data from various languages in the BABEL project, and show a large improvement over systems without tonal features and systems where pitch and POV information was obtained from SAcC or getf0. © 2014 IEEE.

Discover hidden collaborations