Multimedia coding using adaptive regions of interest

The key attributes of neural processing essential to adaptive multimedia processing are presented. The objective is to show why neural networks are a core technology for efficient representation for image information. It is demonstrated how neural network technology gives a solution for extracting a segmentation mask in adaptive region-of-interest (ROI) video coding. The algorithm uses a two-layer neural network architecture that classifies video frames in ROI and non-ROI areas, also being able to adapt its performance automatically to scene changes. The algorithm is incorporated in motion-compensated discrete cosine transform (MC-DCT) based coding schemes, optimally allocating more bits to ROI than to non-ROI areas, achieving better image quality, as well as signal-to-noise ratio improvements compared to standard MPEG MC-DCT encoders.