METHOD AND APPARATUS FOR IMPROVING EFFICIENCY OF IMAGE AND VIDEO CONTENT DELIVERY BASED ON SPATIAL VIEWING DATA

摘要

A method, apparatus and computer program product are provided for improving the efficiency of content delivery based on consumption data. Consumption data may be collected from various users and/or devices describing frequently viewed areas in the content. The consumption data may be represented as a heat map. Variable encoding properties, caching priorities, and/or rendering priorities may be determined based on the consumption data. Content encoding and delivery may therefore be optimized accordingly. Machine learning may be applied to create predicted consumption data for content which lacks actual consumption data, allowing optimizations to be applied in advance of any actual consumption. Some concepts may be applied to unviewed video to optimize live streaming or the like.