This paper proposes a novel semantic communication (SemCom) framework for real-time adaptive bitrate video streaming by integrating the Latent Diffusion Model (LDM) into FFmpeg techniques. To address the high bandwidth usage, storage inefficiency, and QoE degradation associated with conventional CBR and ABR streaming, we compress I-frames into a latent space to achieve storage and semantic transmission savings while maintaining high image quality. B- and P-frames are retained as coordination metadata to enable efficient video reconstruction at the user end. Furthermore, state-of-the-art noise reduction and video frame interpolation (VFI) techniques are integrated to mitigate semantic ambiguity and restore temporal coherence between frames, even in noisy wireless environments. Experimental results demonstrate that the proposed method achieves high-quality video streaming with optimized bandwidth usage and outperforms state-of-the-art solutions in terms of QoE and resource efficiency. This research opens new possibilities for scalable real-time video streaming in 5G and next-generation 5G networks.
