Digital Multimedia Communications by Zhai, Guangtao; Zhou, Jun; Wang, Anhong; Ye, Long;

The two-volume proceedings set constitutes the revised selected papers of the 22nd International Forum of Digital TV and Wireless Multimedia Communications, IFTC 2025, held in Taiyuan, China, during November 20–22, 2025.

The 64 full papers were included in this were carefully reviewed and selected from 165 submissions. They focus to promote technology, equipment, and applications in the field of digital media by comparing the characteristics, frameworks, and significant techniques and their maturity, analyzing the performance of various applications in terms of scalability, manageability, and portability, and discussing the interfaces among varieties of networks and platforms.

.- Affective Computing.

.- Joint Dynamic Electrode Permutation and Hierarchical Fusion for EEG-fNIRS Emotion Classification.

.- Emotion Understanding in Visual Scenes via Instruction-Tuned Multimodal Large Language Models.

.- Deep Learning for Social Media.

.- Intelligent Safety System for Coal Mines Based on 3D Human Reconstruction Using mmWave Radar.

.- MC-AVD: Generalizable Audio-Visual Deepfake Detection via Multi-Modal Collaboration.

.- Gradient Distribution Balancing for Dual Imbalanced Long-Tailed Learning.

.- Comparative Study on Feature Selection for Intrusion Detection Based on PPO and GRPO Algorithms.

.- Graphics and Image Processing for Virtual Reality.

.- Enhancing the Power of GUI Agents by Reinforcement Learning.

.- AmoMimic: Speech-Driven 3D Face Based on Emotion Enhancement and Multimodal Approach.

.- Implicit Neural Representation with Structured Sampling and Hybrid Position Encoding for Point Cloud Compression.

.- Ultra-Efficient Same-Accuracy Mapping: Non-Uniform B-Splines with Incremental Voxel Acceleration.

.- Enhanced Low-Bitrate Light Field Video Compression Based on Key Sparse View Sequence Coding.

.- Large Language Models.

.- Mitigating Proactive and Retroactive Interference in LLMs: A Training-Free Framework for Robust Contextual Working Memory.

.- INGF: Integrated News Generation Framework with Multimodal Editing and Alignment.

.- CLGA: A Linguistic Arena for Testing LLMs’ Prowess in Chinese Literary Games.

.- Hybrid Large-Small Model Framework for Core Business Consistency Evaluation.

.- Multimodal Auxiliary Analysis System for Oral Disease Diagnosis Based on Fine-Tuned Models: Diagnosis and Quality Evaluation.

.- VidDPOK: Aligning Text-to-Video Generation with Human Preferences via Reinforcement Learning.

.- Media Quality Assessment.

.- Genre-Aware Representation Learning for No-Reference Game Video Quality Assessment.

.- Brain-Inspired Audio Quality Assessment Based on Audio-EEG Feature Fusion.

.- Multimedia Applications and Technologies.

.- Decoupled Prototype Network Based on Grassmann Manifold Orthogonal Projection.

.- GP-FusionNet: Accurate Depth-Guided 3D Reconstruction of Textureless Scenes with Gaussian Splatting.

.- Fine-Grained Respiratory Waveform Reconstruction with mmWave Radar by VMD and Lightweight CNN-BiGRU-AT Hybrid Network.

.- Hearing the Scene: Contextual Effects of Spatial Audio Elements in VR.

.- SynSinger: Audio-driven Singing Avatar Motion Synthesis.

.- Dynamic Aperture Expansion for UAV Swarm-Based Phased-Array Radar Systems.

.- Prototype Similarity-based Adaptive Clustered Federated Learning Framework.

.- BotSphere: A Multi-Agent Framework for Modeling Social Bot Infuence on Public Opinion Dynamics.

.- Performance Analysis of Hybrid RIS Assisted Full Duplex Systems with LMMSE Channel Estimation.

.- Weight-Aware Synergy of Deep Reinforcement Learning and Evolutionary Algorithms for Multi-Objective VRPTW.

.- SCMA-OAC Enabled Federated Learning.

.- Research on Intelligent Monitoring and Early-Warning Technology for Clearance Safety in Power Transmission Line Construction.