Delayed suppression normalizes face identity responses in the primate brain

SummaryThe primate brain is specialized for social visual perception. Previous work indicates that recognition draws upon an internal comparison between a viewed face and an internally stored average face. Here we demonstrate that this comparison takes the form of a delayed, dynamic suppression of face averageness among single neurons. In three macaque face patches, spiking responses to low-identity morphed faces met with a synchronous attenuation starting approximately 200 ms after onset. Analysis showed that a late-emerging V-shaped identity tuning was sometimes superimposed on linear ramp tuning. This pattern could not be ascribed to repetition suppression within a given session. The results indicate that the brain’s analysis of faces is enhanced through predictive normalization of identity, which increases sensitivity among face-selective neurons to distinctive facial features known to drive recognition.

Efficient Design of ADC BIST with an Analog Ramp Signal Generation and Digital Error Estimation

This paper presents the design of linear ramp generator and digital BIST for an on-chip ADC testing. It replaces the costly and time-consuming traditional mixed signal test methods like DSP-based testing, ATE, etc. The proposed on-chip analog ramp generator uses only a few transistors to generate linear ramp signal. A TIQ comparator based 8-bit flash ADC is taken under test. The output response of the ADC is analyzed in the digital BIST to measure the primary nonidealities affecting the linearity and accuracy of the data conversion. In testing, ADC generates the digital data sequence as a test pattern in response to the ramp input while digital BIST estimates the conversion error. This method does not require DAC and any additional components which increase the area overhead of ADC test. The complete design of ramp generator is integrated with TIQ flash ADC and verified in 0.18[Formula: see text][Formula: see text]m CMOS technology with 1.8[Formula: see text]V of the power supply and 100[Formula: see text]kHz of the input frequency. Measurement of nonidealities shows that the design of an 8-bit flash ADC has good accuracy in data conversion with the differential nonlinearity of [Formula: see text]0.24/[Formula: see text]0.17[Formula: see text]LSB and integral nonlinearity of [Formula: see text]0.44/[Formula: see text]0.04[Formula: see text]LSB.

Investigation of Wave Overtopping from Various Geometrical Configurations of Coastal Breakwater Ramps

Breakwater is often required for erosion protection. The effectiveness of breakwater structure as a coastal defence could be determined through overtopping wave surpass onto its crest freeboard structure. The common design practice of breakwater structure is using a linear ramp configuration. There is still a limited knowledge on the contribution of various configuration ramps shape into breakwater efficiency. Thus, this study attempts to investigate a few other's configuration of ramp geometries through overtopping discharge rate performance and identified its relationships with the crest freeboard. Three configurations of ramp geometries (convex, concave and cubic) have been investigated using numerical simulation of CFD-Flow 3D software. The result indicates that, there are significant influences of ramp configuration parameters on the overtopping rate and the highest overtopping rate is contributed from cubic shape ramp geometry. It is recommended that the detail study on cubic ramp shape geometry for determining its capability in increasing overtopping rate.

AN EFFICIENT PARTIAL DOMAIN ENRICHED ELEMENT-FREE GALERKIN METHOD CRITERION FOR CRACKS IN NONCONVEX DOMAINS

In the present work, an efficient partial domain, intrinsic, enriched element-free Galerkin criterion has been extended to simulate the cracks lying in nonconvex domains. According to this criterion, only a part of the domain near the crack tip has been enriched. A linear ramp function has been used to avoid the sudden truncation of the enrichment effect. Some cases of cracks lying in convex as well as in nonconvex domains have been solved by both full and partial domain enrichment criteria under plane stress conditions. For the cracks lying in convex domain, the results obtained by full domain enrichment criterion are found in good agreement with those obtained by partial domain enrichment criterion, whereas for cracks lying in nonconvex domain, the results obtained by full domain enrichment criterion are found to be misleading. The partial domain enrichment not only accurately simulates the cracks in nonconvex domains but also reduces the computational cost of the method.