DSP-based neural systems for the perceptual assessment of visual quality

The paper presents an efficient hardware realization of circular back propagation (CBP) networks on digital signal processors (DSP). The resulting neural-system is aimed at enhancing "smart" TV displays and supports the estimation of perceptual quality of visual signals. The DSP-based neuro-platform operates on raw digital signals and yields the associate estimate of perceived quality in real time. Experimental results confirm both evaluation accuracy and timing performance.