Approximate functional testing for image applications based on error-tolerance
| Autor: | Tai-Ang Cheng, Chao-Ru Chen, Tong-Yu Hsieh |
|---|---|
| Rok vydání: | 2017 |
| Předmět: |
business.industry
Computer science Functional testing 02 engineering and technology Machine learning computer.software_genre Reliability engineering Software Error tolerance Logic gate 0202 electrical engineering electronic engineering information engineering Dependability 020201 artificial intelligence & image processing Test requirements Artificial intelligence business computer |
| Zdroj: | 2017 IEEE International Conference on Consumer Electronics - Taiwan (ICCE-TW). |
| DOI: | 10.1109/icce-china.2017.7991066 |
| Popis: | Error-tolerance is a notion that focuses on evaluating the acceptability of errors by considering the insensitivity of human beings to minor vibrations in multimedia applications. In this paper we will show that this notion can relax the functional test requirements of a target circuit. Rather than finely grading the quality of the output results as conventional methods do, we only need to decide if the circuit is acceptable or not. We refer to this notion as approximate functional testing in this paper. This notion can lead to great reduction in test time and thereby the test cost. Cost-effective hardware implementation of an on-chip quality monitor circuit can also thus be enabled for evaluating the dependability of the circuit. In this paper a flow will be presented to describe how to carry out approximate testing. A case study on image applications will also be provided to illustrate the flow. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|