ICASSP 2004

318 papers

Top Research Directions

Research directions most represented by ICASSP 2004 papers.

Constrained Coding for DNA Storage and Synthesis 1
Resolving Front-Back Ambiguity in Acoustic Localization 1
Direct Minimization of Classification Error 1
Scaling Noisy Data for Better Generalization 1
Adaptive Thresholding for Dynamic Anomaly Detection 1
Joint Latent State and Parameter Estimation 1
Aligning Sound Sources Across Frequency Bins 1
Bootstrapping Speech Tasks with Pretrained Acoustic Models 1
Speech Transformation Without Parallel Training Data 1
Adapting Language Models to Temporal Drift 1
Convolutive Blind Source Separation in Reverberant Environments 1
Robust Audio Modeling Under Acoustic Domain Shift 1

Top Authors

AuthorPapers
Alfred O. Hero8
Georgios B. Giannakis5
Hiroshi Sawada5
Rama Chellappa5
Ananthram Swami4
Deniz Erdogmus4
Phil Woodland4
Wing-Kin Ma4
Brian M. Sadler3
Jeff Bilmes3
Konstantinos N. Plataniotis3
Lawrence K. Saul3
Ramani Duraiswami3
Samy Bengio3
Shoko Araki3
S Kevin Zhou3
Alan Wee-Chung Liew2
Alper Tunga Erdogan2
Christian Fuegen2
Dilek Hakkani Tur2

Impactful Papers

Papers from ICASSP 2004 that shaped research clusters.

Foundational

Influential

Papers

Title
Towards language portability in statistical speech translation
Towards optimal quantization in multistage audio coding
Toward sound-based synthesis: the far-field case
Towards practical deployment of audio-visual speech recognition
Tracking multiple speakers using random sets
Trapping conversational speech: extending TRAP/tandem approaches to conversational telephone speech recognition
Underdetermined blind separation for speech in real environments with sparseness and ICA
Unified framework for a class of frequency-offset estimation techniques for OFDM
Unified framework for a class of frequency-offset estimation techniques for OFDM.
Unsupervised and active learning in automatic speech recognition for call classification
Unsupervised learning of sparse and shift-invariant decompositions of polyphonic music
Using linear prediction to enhance the tracking of partials [musical audio processing]
Vehicle detection and tracking using acoustic and video sensors
Vocabulary-independent search in spontaneous speech
Wavelet and footprint sampling of signals with a finite rate of innovation
Wavelet-based denoising by customized thresholding
Weighted motion estimation for efficiently coding scene transition video
Why do multi-stream, multi-band and multi-modal approaches work on biometric user authentication tasks?

« prev Page 7/7