2011 Digital Signal Processing and Signal Processing Education Meeting (DSP/SPE) 2011
DOI: 10.1109/dsp-spe.2011.5739225
|View full text |Cite
|
Sign up to set email alerts
|

Quantization for classification accuracy in high-rate quantizers

Abstract: Quantization of signals is required for many transmission, storage and compression applications. The original signal is quantized at the encoder side. At the decoder side, a replica of the original signal that should resemble the original signal in some sense is recovered. Present quantizers make an effort to reduce the distortion of the signal in the sense of reproduction fidelity. Consider scenarios in which signals are generated from multiple classes. The encoder focuses on the task of quantizing the data w… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1

Citation Types

0
1
0

Year Published

2014
2014
2019
2019

Publication Types

Select...
2
1
1

Relationship

0
4

Authors

Journals

citations
Cited by 4 publications
(1 citation statement)
references
References 10 publications
0
1
0
Order By: Relevance
“…It was shown recently that considering the application and designing data codecs appropriately, that do not maximize fidelity type criteria (such as the mean squared error), but consider how would an analysis algorithm (e.g., a classifier) perform on compressed data, is beneficial from a bit rate perspective [17]. This notion was explored in [17] and [18] with respect to quantization, however, as of now the design of color transforms optimized particularly for classification accuracy has not been considered yet. Thus, given some previously labeled data, we propose a methodology to obtain application-dependent color transforms, that while aiming to retain energy compaction properties, also try to maximize separability of the transformed data.…”
Section: Introductionmentioning
confidence: 99%
“…It was shown recently that considering the application and designing data codecs appropriately, that do not maximize fidelity type criteria (such as the mean squared error), but consider how would an analysis algorithm (e.g., a classifier) perform on compressed data, is beneficial from a bit rate perspective [17]. This notion was explored in [17] and [18] with respect to quantization, however, as of now the design of color transforms optimized particularly for classification accuracy has not been considered yet. Thus, given some previously labeled data, we propose a methodology to obtain application-dependent color transforms, that while aiming to retain energy compaction properties, also try to maximize separability of the transformed data.…”
Section: Introductionmentioning
confidence: 99%