Applications -- A Proxemic Multimedia Interaction over the Internet of Things -- Outdoor Air Quality Inference from Single Image -- Multimodal Music Mood Classification by Fusion of Audio and Lyrics -- Multidimensional Context Awareness in Mobile Devices -- AttRel: An Approach To Person Re-Identification By Exploiting Attribute Relationships -- Sparsity-based Occlusion Handling Method for Person Re-identification -- Visual attention driven by auditory cues -- A Synchronization Ground Truth for the Jiku Mobile Video Dataset -- Mobile Image Analysis: Android vs. iOS -- Dynamic User Authentication Based on Mouse Movements Curves -- Sliders Versus Storyboards - Investigating Interaction Design for Mobile Video Browsing -- Performance Evaluation of Students using Multimodal Learning Systems -- Is your first impression reliable? Trustworthy analysis using facial traits in portraits -- Wifbs: A Web-based Image Feature Benchmark System.-Personality Modeling based Image Recommendation -- Aesthetic QR Codes Based on Two-Stage Image Blending -- Person Re-identification Using Data-driven Metric Adaptation.-Factorizing Time-Aware Multi-Way Tensors for Enhancing Semantic Wearable Sensing -- User-Centred Evaluation to Interface Design of E-Books -- A New Image Decomposition and Reconstruction Approach -- Adaptive Fourier Decomposition -- Video Showcase Graph-Based Browsing for Large Video Collections -- Enhanced Signature-based Video Browser VERGE: A Multimodal Interactive Video Search Engine -- IMOTION - a Content-based Video Retrieval Engine -- A Storyboard-based Interface for Mobile Video Browsing -- Collaborative Browsing and Search in Video Archives with Mobile Clients.-The Multi-Stripe Video Browser for Tablets -- NII-UIT Browser: A Multimodal Video Search System -- Interactive Known-Item Search Using Semantic Textual and Colour Modalities -- Demonstration -- ImageMap - Visually Browsing Millions of Images -- Dynamic Hierarchical Visualization of Keyframes in Endoscopic Video Facial AgingSimulator by Data-driven Component-based Texture Cloning -- Affective Music Recommendation System based on the Mood of Input Video -- MemLog, an Enhanced Lifelog Annotation and Search Tool -- Software Solution for HEVC Encoding and Decoding -- A Surveillance Video Index and Browsing System Based on Object Flags and Video Synopsis A Web Portal For Effective Multi-model Exploration -- Wearable Cameras for Real-time Activity Annotation -- Personal (Big) Data Modeling for Information Access & Retrieval -- Making Lifelogging Usable: Design Guidelines for Activity Trackers -- Towards Consent-Based Lifelogging in Sport Analytic -- A Multi-Dimensional Data Model for Personal Photo Browsing -- Discriminative Regions: A Substrate for Analyzing Life-logging Image Sequences -- Fast Human Activity Recognition in Lifelogging -- Social Geo-Media Analytics and Retrieval -- Iron Maiden while jogging, Debussy for dinner? -- Travel Recommendation via Author Topic Model based Collaborative Filtering -- Robust User Community-aware Landmark Photo Retrieval -- Cross -domain Concept Detection with Dictionary Coherence by Leveraging Web Images -- Semantic correlation mining between images and texts with global semantics and local mapping -- Image Taken Place Estimation via Geometric Constrained Spatial Layer Matching -- Image or video processing, semantic analysis, and understanding -- Recognition of meaningful human actions for video annotation using EEG based user responses -- Challenging Issues in Visual Information Understanding Researches -- Emotional Tone-Based Audio Continuous Emotion Recognition A Computationally Efficient Algorithm for Large Scale Near-Duplicate Video Detection SLOREV: Using Classical CAD Techniques for 3D Object Extraction from Single Photo Hessian regularized sparse coding for human action recognition -- Robust Multi-label Image Classification with Semi- Supervised Learning and Active Learning -- Photo Quality Assessment with DCNN that Understands Image Well -- Non-negative Low-rankand Group-sparse Matrix Factorization -- Two-Dimensional Euler PCA for Face Recognition Multiclass Boosting Framework for Multimodal Data Analysis.

The two-volume set LNCS 8935 and 8936 constitutes the thoroughly refereed proceedings of the 21st International Conference on Multimedia Modeling, MMM 2015, held in Sydney, Australia, in January 2015. The 49 revised regular papers, 24 poster presentations,  were carefully reviewed and selected from 189 submissions. For the three special session, a total of 18 papers were accepted for MMM 2015. The three special sessions are Personal (Big) Data Modeling for Information Access and Retrieval, Social Geo-Media Analytics and Retrieval, and Image or video processing, semantic analysis and understanding. In addition, 9 demonstrations and 9 video showcase papers were accepted for MMM 2015. The accepted contributions  included in these two volumes  represent the state-of-the-art in multimedia modeling research and cover a diverse range of topics including: Image and Video Processing, Multimedia encoding and streaming, applications of multimedia modelling and 3D and augmented reality.