MultiMedia Modeling

This five-volume set LNCS 15520-15524 constitutes the proceedings of the 31st International Conference on Multimedia Modeling, MMM 2025, held in Nara, Japan, January 810, 2025.
The 135 full papers and 41 short papers presented in these proceedings were carefully reviewed and selected from 348 submissions. The MMM conference was organized in topics related to multimedia modelling, particularly: audio, image, video processing, coding and compression; multimodal analysis for retrieval applications, and multimedia fusion methods.

Inhalt

Regular Papers.- A Dual-Branch Model for Color Constancy.- A Multi-Aspect Multi-Granularity Pronunciation Assessment Method Based on Branchformer Encoder and Hierarchical Aggregation.- A Multi-Expert Collaborative Framework for Multimodal Named Entity Recognition.- A Novel Human Abnormal Posture Detection Method Based on Spatial-Topological Feature Fusion of Skeleton.- AD2AT: Audio Description to Alternative Text, a Dataset of Alternative Text from Movies.- AMFT-YOLO: A Adaptive Multi-Scale YOLO Algorithm with Multi-Level Feature Fusion for Object Detection in UAV Scenes.- AMPLE: Emotion-Aware Multimodal Fusion Prompt Learning for Fake News Detection.- An Analytical Method for Rendering Plenoptic Cameras 2.0 on 3D Multi-Layer Displays.- Balancing Efficiency and Accuracy: An Analysis of Sampling for Video Copy Detection.- BiCA-YOLO: Bidirectional Feature Enhancement and Cross Coordinate Attention for Small Object Detection.- BLCC: A Benchmark for Multi-LiDAR and Multi-Camera Calibration.-Boosting Human Pose Estimation via Heatmap Refinement.- Camouflaged Object Detection Based on Localization Guidance and Multi-Scale Refinement.- Chain of Thought Guided Few-shot Fine-tuning of LLMs for Multimodal Aspect-based Sentiment Classification.- CLIP Multi-modal Hashing for Multimedia Retrieval.- Comparative Analysis of Relevance Feedback Techniques for Image Retrieval.- Cross-View Geo-Localization via Learning Correspondence Semantic Similarity Knowledge.- DART: Depth-Enhanced Accurate and Real-Time Background Matting. Data-free Functional Projection of Large Language Models onto Social Media Tagging Domain.- Deep Dual Internal Learning for Hyperspectral Image Super-Resolution.- Detoxification of Unlabeled Dataset: Reducing Implicit Class Imbalance Using Pseudo-Jacobian of GAN's Generator.- DistillSleep: Leverage Self-Distillation to Improve Performance After Representation Learning for Sleep Staging.- DocMamba: Robust Document Image Dewarping via Selective State Space Sequence Modeling.- Dual-Task Feedback Learning for Tongue Detection via Super-Resolution Integration.- Dynamic Exploration Graph: A Novel Approach for Efficient Nearest Neighbor Search in Evolving Multimedia Datasets.- EIA: Edge-aware Imperceptible Adversarial Attacks on 3D Point Clouds.- Enhancing Environmental Monitoring through Multispectral Imaging: The WasteMS Dataset for Semantic Segmentation of Lakeside Waste.- ESC-MISR: Enhancing Spatial Correlations for Multi-Image Super-Resolution in Remote Sensing.- Flat Local Minima for Continual learning on Semantic Segmentation.- FoodMLLM-JP: Leveraging Multimodal Large Language Models for Japanese Recipe Generation.- Frequency-aware Convolution for Sound Event Detection.- Frequency-Based Unsupervised Low-Light Image Enhancement Framework.- GFA-UDIS: Global-to-Flow Alignment for Unsupervised Deep Image Stitching.