Spatio-Temporal Indexing, Content-Aware Compression, and Adaptive Delivery for Bandwidth-Efficient Live Video Streaming: A State-of-the-Art Review

Abstract

Live video streaming accounts for a substantial share of contemporary internet traffic; however, efficiently utilizing constrained bandwidth while preserving temporal continuity and perceptual quality remains a significant challenge. Addressing this challenge requires efficient video compression, timely delivery, and adaptive quality management under varying content complexity and network conditions. Conventional video codecs exploit spatial and temporal redundancies to reduce bitrate, while adaptive bitrate (ABR) streaming systems dynamically select encoded representations based on estimated network capacity and playback conditions. similarly, keyframe extraction and video indexing identify representative frames, scene transitions, and temporal structures to facilitate content retrieval and analysis. Despite their complementary capabilities, these approaches address distinct objectives, and their integration into a unified live-streaming architecture introduces challenges related to computational overhead, end-to-end latency, coding dependencies, and temporal continuity. This paper presents a state-of-the-art review of established video coding standards, content-aware frame selection, spatio-temporal indexing, compression-domain analysis, and adaptive live-streaming architectures. Based on a critical synthesis of the literature, it identifies unresolved research gaps and highlights promising directions, including lightweight content-aware analysis, dependency-aware temporal selection, joint rate–distortion–latency optimization, and cross-layer feedback between encoding and delivery. The review provides a conceptual foundation for future research into integrated, bandwidth-efficient live video streaming without assuming that the proposed directions have been experimentally validated.