ASAP 2011 - 22nd IEEE International Conference on Application-Specific Systems, Architectures and Processors 2011
DOI: 10.1109/asap.2011.6043265
|View full text |Cite
|
Sign up to set email alerts
|

Instruction set extension for high throughput disparity estimation in stereo image processing

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
1

Citation Types

0
4
0

Year Published

2012
2012
2018
2018

Publication Types

Select...
3
3
1

Relationship

0
7

Authors

Journals

citations
Cited by 12 publications
(4 citation statements)
references
References 14 publications
0
4
0
Order By: Relevance
“…Typically, soft CPU customization is undertaken with the target of accelerating the most frequently used functions. This has resulted in performance improvements as reported in previous publications [5,6]. ISEs usually target applications where the CPU time is dominated by smaller code sections (i.e.…”
Section: Introductionmentioning
confidence: 61%
“…Typically, soft CPU customization is undertaken with the target of accelerating the most frequently used functions. This has resulted in performance improvements as reported in previous publications [5,6]. ISEs usually target applications where the CPU time is dominated by smaller code sections (i.e.…”
Section: Introductionmentioning
confidence: 61%
“…The number of index bits depends on the value set in the configuration register. In one cycle, the LFSR state is updated by the same number of LFSR steps as index bits used (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16). Additionally, the instruction set extension (ISE) is realized as a multi-cycle instruction, which allows handling of one sample in a number of cycles equal to the configured number of non-zero entries (I) per matrix column.…”
Section: Configurabilitymentioning
confidence: 99%
“…Application-specific instruction-set processors (ASIPs) can compensate for the performance degradation issue, since they are optimized for a specific application domain, providing increased efficiency and performance for the core algorithms of the domain's target applications. For instance, an ASIP optimized for stereo image processing can achieve up to 130× speed-up compared to a conventional processor [11]. These performance optimizations also lead to energy saving as in [12], where a processing core with few accelerators dedicated to biomedical applications, can achieve up to 11.5× energy saving compared the processing core-only implementation.…”
Section: Introductionmentioning
confidence: 99%
“…For instance, an ASIP optimized for stereo image processing can achieve up to 130x speed-up compared to a conventional processor [10]. These performance optimizations also lead to energy saving as in [11], where a processing core with few accelerators dedicated to biomedical applications, can achieve up to 11.5x energy saving compared the processing core-only implementation.…”
Section: Introductionmentioning
confidence: 99%