Search

User adaptation is critical in the future design of human-computer interaction systems. Many pattern recognition problems, such as handwriting/sketching recognition and speech recognition, are user dependent since different users' handwritings, drawing styles, and accents are different. Hence, the classifiers for solving these problems should provide the functionality of user adaptation so as to let all users experience better recognition results. However, the user adaptation functionality requires the classifiers have the incremental learning ability in order to learn fast. In this paper, an SVM-based incremental active learning algorithm is presented to solve this problem. By utilizing the support vectors and only a small portion of the nonsupport vectors as well, in addition to the new interrogative samples, in the iterative training and reclassification cycle, both the training time and the storage space are saved with only very little classification precision being lost. Theoretical analysis, experimentation, evaluation, and real application samples in our online graphics recognition system are presented to show the effectiveness of this algorithm.

A special calibration tool has been developed for a CCD camera based vision system in an automatic assembly machine. The machine is used to attach orifice plates onto a silicon wafer in a production process. The center locations of the positioning circular holes on the plate must be controlled accurately to coincide with those on the wafer die before they are attached together by UV curing. Although CCD camera based vision systems are widely used for accurate positioning and dimensional measurements in precision engineering, electronics and semiconductor industry, their calibrations are normally done by artefacts with plane patterns. These artefacts are therefore restricted to only two dimensional measurements. The calibration tool we developed was to check the positioning accuracy of circular objects in a two-layered structure. It can also be used to determine parallax errors, non-linearity and spatial non- uniformity errors as well as repeatability of the vision system with an uncertainty at sub-micrometer level. The design, calibration and performance of the tool are described in detail in this paper.

We have constructed a laser measurement system for automatic calibration of height setting micrometer. This automation is accomplished by an interactive software tool which performs essential calibration functions. The instrument incorporates a double frequency stabilized laser based on the heterodyne measurement technique. Positioning of height setting micrometer is undertaken by a X-Y stage driven by the servo motors and calibration is performed by combining laser and electronic probe readings. Abbe offset tilting error is eliminated by using the double-path interferometer scheme. The instrument has a capability of measurement range 0 - 350 mm, resolution 40 nm, expanded uncertainty of height measurement better than 0.5 micrometer at the 95% confidence level over the full range. The whole calibration procedure can be completed in two hours making it much more efficient than the conventional method of gauge block comparison. The instrument can also be applied to various linear measurements with high accuracy to reduce the measuring time and to simplify the measurement setup.

A method is described for the calibration of vertical piezoelectric transducer (PZT) stages using a large range metrological atomic force microscope (LRM-AFM). A vertical PZT stage is mounted onto the system andanoptical-flat sampleisattachedonthetopofthePZTstage.TheAFM probe is operated in contact mode and used as a null indicator to measure the movement of the optical-flat as it is driven by the PZT stage. As the PZT stageisscannedvertically,theAFMprobeisincontactwiththetestsurface without lateral scanning. The displacement of the vertical stage of the LRM-AFM is measured by the integrated laser interferometer, and the corresponding laser interferometer readings and the nominal position of the PZT stage are recorded. This body of data is then used to establish the relationship between the laser interferometer reading and the nominal displacement of the PZT stage. Our results show that this method can be used to calibrate PZT stages up to a range of 250 μm with an expanded uncertainty of less than 5 nm. © 2012 Society of Photo-Optical Instrumentation Engi

Following the decisions of the 1994 meeting of the Consultative Committee for Photometry and Radiometry (CCPR), BIPM acted as pilot laboratory for an international comparison of cryogenic radiometers, later designated as a supplementary comparison CCPR-S3, carried out using silicon trap detectors as transfer devices. This comparison was completed in February 1999. After the completion of this comparison and the publication of the final report, the National Metrology Centre of the Agency for Science, Technology and Research (NMC-A*STAR) of Singapore expressed a wish to demonstrate its primary spectral responsivity measurement capabilities, in support of its CMC claims, through linkage to the CCPR-S3 comparison. NPL offered to organize a bilateral comparison with NMC-A*STAR following a similar protocol to that used in the original comparison. The comparison was carried out by using three silicon trap detectors as transfer devices, to be calibrated at a number of laser wavelengths at NPL and NMC-A*STAR from 2005 to 2007. The results of the comparison demonstrate agreement between NMC-A*STAR and CCPR-S3 comparison reference values, in the selected wavelength range, to better than 6.8 parts in 104 with a combined uncertainty (k = 1) of 2.2 parts in 104 or better. Main text. To reach the main text of this paper, click on Final Report. Note that this text is that which appears in Appendix B of the BIPM key comparison database kcdb.bipm.org/. The final report has been peer-reviewed and approved for publication by the CCPR, according to the provisions of the CIPM Mutual Recognition Arrangement (MRA).

APMP.PR-S1, a supplementary comparison of irradiance responsivity of UVA detectors, was carried out among seven national metrology institutes piloted by SPRING Singapore from 2003 to 2005. Two quantities, the narrow band UV (365 nm ± 5 nm) irradiance responsivity and the broad band UVA (315 nm–400 nm) irradiance responsivity of the transfer detectors, have been compared. Commercial UV source (medium pressure mercury short arc lamp) and UVA detectors were used as transfer standards in the comparison. Measurement results from participants were reported and their uncertainties associated with the comparison were analysed in this report. The method of weighted mean with cut-off was used to calculate the comparison reference values. The results from most participating labs lie within ±5% against the comparison reference values with a few exceptions. The degree of agreement of the comparison depends not only on the base scales of spectral responsivity and spectral irradiance of a laboratory, but also equally importantly on the method used for the measurement. Main text. To reach the main text of this paper, click on Final Report. Note that this text is that which appears in Appendix B of the BIPM key comparison database kcdb.bipm.org/. The final report has been peer-reviewed and approved for publication by the APMP, according to the provisions of the CIPM Mutual Recognition Arrangement (MRA).

A bilateral comparison of spectral irradiance using tungsten halogen lamps as transfer standards covering the spectral region from 250 nm to 1600 nm was carried out between the NMIA and the SPRING Singapore in the period from November 2003 to August 2004. The NMIA acted as the pilot laboratory and provided the link to the key comparison CCPR-K1.a. The spectral irradiance units measured by the NMIA and the SPRING Singapore in this comparison are in agreement within the combined standard uncertainties of the laboratories over the wavelength range compared. The SPRING Singapore's spectral irradiance scale is based on a group of four NIST-calibrated standard lamps, three of which were calibrated prior to 2001 and one of which was calibrated in late 2001 after NIST adopted a new detector-based scale. There appear to be systematic differences between the two scales which are close to differences between the NMIA scale and the two NIST scales as represented by the group of reference lamps used at the SPRING Singapore. Main text. To reach the main text of this paper, click on Final Report. Note that this text is that which appears in Appendix B of the BIPM key comparison database kcdb.bipm.org/. The final report has been peer-reviewed and approved for publication by the CCPR, according to the provisions of the CIPM Mutual Recognition Arrangement (MRA).

The observation of self-induced transparency on the magnetic dipole 2P1/2 F = 3 5 2P3/2 F′ = 4, j MF = 0 transition of atomic iodine has been extended to cover pulse areas up to 4 ~. Laser pulses of various areas have been shown to be reshaped considerably by the coherent field-matter interaction. Pulses have been found to narrow by two-thirds for a 3 ~ -pulse. The lower limit of the homogeneous relaxation time T′2 of atomic iodine vapour is determined from pulse broadening and transmission measurements.

Coherent propagation of intense short laser pulses through degenerate absorbing media is investigated with the use of an atomic-iodine laser-absorber self-resonant combination. Four degenerate systems, the $^{2}P_{\frac{1}{2}}\ensuremath{-}^{2}P_{\frac{3}{2}}$, $F=3\ensuremath{\leftrightarrow}{F}^{\ensuremath{'}}=4$, $\ensuremath{\Delta}{M}_{F}=0,\ifmmode\pm\else\textpm\fi{}1$ and $F=2\ensuremath{\leftrightarrow}{F}^{\ensuremath{'}}=2$, $\ensuremath{\Delta}{M}_{F}=0,\ifmmode\pm\else\textpm\fi{}1$ transitions, are studied under various conditions. Theoretical analysis based on the "pulse-area---pulse-energy" approach shows different pulse propagation behaviors for three typical types of degeneracy. Experimental results give good agreement with theoretical predictions. It is concluded that self-induced transparency exists in any degenerate two-level system, provided that suitable polarization of radiation is used. The usefulness of self-induced-transparency phenomena for measurements of transitional dipole moment and homogeneous relaxation time is also demonstrated.

Self-induced transparency (SIT) has been observed on a Q(J)-branch transition (J=2) for the first time. Coherent field-matter interaction was studied using a photodissociation atomic iodine laser system operating on the 2P1/2 F=2 →2P3/2 F′=2 magnetic dipole hyperfine transition and a self-resonant iodine vapour. With linearly polarized radiation the propagation phenomena are characterized by high transparency and large time delay at pulse areas θ > π and small-pulse splitting at θ ≃ π. With circularly polarized radiation, non-degenerate SIT is observed. The results can be explained by the dynamic effects of the different dipole moments involved in the process.

A multi-functional differential spectral responsivity (DSR) measurement system for primary calibration of the shortcircuit current (SCC) of reference solar cells with WPVS (World Photovoltaic Scale) design [1] is reported in this paper. Based on the dual beam principle, the system is able to measure DSR at different bias levels, spatial uniformity, linearity, temperature coefficient of the spectral responsivity of reference cells in the spectral range from 280 nm to 1200 nm. The measurement principle, system design and construction as well as the performance of the system are described in this work. An uncertainty ≤ 1% for SCC measurement is confirmed by experimental results.