ICME 2018

126 papers
The 2018 IEEE International Conference on Multimedia and Expo (ICME) featured 179 papers focusing on the intersection of human behavior and multimedia processing. Prominent research themes included the application of human memory models to video quality assessment and the modeling of stochastic visual attention. Additionally, the conference highlighted developments in computational psycholinguistics and personality recognition from digital footprints, indicating a shift toward behavioral prediction through multimedia data. Technical contributions emphasized visual tracking, spatial-temporal optimization, and neural network efficiency. Specific advancements included adaptive layerwise quantization for deep neural network compression and the specialization of networks for individual video sequences. Other research explored unsupervised representation learning through temporal slowness and cross-view geo-localization. Application-focused studies addressed topics such as cascaded human parsing, self-supervised road segmentation, and neural network-based inter prediction for High Efficiency Video Coding (HEVC). Prolific authors for this edition included Yaowei Wang, Yonghong Tian, and Houqiang Li, among others.

Top Research Directions

Research directions most represented by ICME 2018 papers.

Efficient Feature and Search Management for Real-Time Tracking 2
Repurposing GAN Discriminators for Representation Learning 1
Reconstructive Regularization for Information Preservation in Latent Representations 1
Hierarchical Vocabulary Factorization for Efficient Language Models 1
Why Weight Magnitude Fails as a Pruning Heuristic 1
Probabilistic Modeling of Human Gaze and Scanpaths 1
Balancing Task Discrimination and Domain Invariance 1
Replacing Fully Connected Layers with Global Average Pooling 1
Scaling Psychological Assessment via Digital Footprints 1
Selective Spatial Computation for Efficient Vision 1
Alternating Optimization for Low Rank Tensor Decompositions 1
Tracking Input Coverage in Attention Mechanisms 1
Phoneme Specific Spectral Modulation for Vocal Emotion 1
Learning Invariant Representations via Temporal Slowness 1
Compositional Generalization through Primitive Decomposition 1

Top Authors

AuthorPapers
Houqiang Li4
Wei Feng4
Wengang Zhou4
Yemin Shi4
Yonghong Tian4
Lizhuang Ma3
Qinghua Hu3
Qingming Huang3
Ruimin Hu3
Wen Gao3
Xiaojie Guo3
Bingpeng Ma2
Ce Zhu2
Changqing Zhang2
Debin Zhao2
Fanyang Meng2
Fei Su2
Hong Liu2
Jian Guan2
Jia Wu2

Impactful Papers

Papers from ICME 2018 that shaped research clusters.

Influential

Papers

Title
Seethevoice: Learning from Music to Visual Storytelling of Shots
Semantic Manifold Alignment in Visual Feature Space for Zero-Shot Learning
SFCM: Learn a Pooling Kernel for Weakly Supervised Object Localization
Simultaneous Realization of Multiple Music Video Applications Based on Heterogeneous Network Analysis Via Latent Link Estimation
Single Image Layer Separation via Deep Admm Unrolling
Skeleton-Based Human Action Recognition Using Spatial Temporal 3D Convolutional Neural Networks
Skeleton-Indexed Deep Multi-Modal Feature Learning for High Performance Human Action Recognition
Soft Clustering Guided Image Smoothing
Sparse Representation for Color Image Based on Geometric Algebra
Spatio-Temporal Large Margin Nearest Neighbor (ST-LMNN) Based on Riemannian Features for Individual Identification
Spherical Structural Similarity Index for Objective Omnidirectional Video Quality Assessment
Structure-Texture Decomposition via Joint Structure Discovery and Texture Smoothing
Super-Resolution Quality Assessment: Subjective Evaluation Database and Quality Index Based on Perceptual Structure Measurement
Support Vector Metric Learning on Symmetric Positive Definite Manifold
Temporal Attentive Network for Action Recognition
Tensor Sensing for Rf Tomographic Imaging
TLR: Transfer Latent Representation for Unsupervised Domain Adaptation
Towards Compact Visual Descriptor via Deep Fisher Network with Binary Embedding
Trajectory Factory: Tracklet Cleaving and Re-Connection by Deep Siamese Bi-GRU for Multiple Object Tracking
Transformation on Computer-Generated Facial Image to Avoid Detection by Spoofing Detector
Transim: Transfer Image Local Statistics Across EOTFS for HDR Image Applications
Unsupervised Local Facial Attributes Transfer Using Dual Discriminative Adversarial Networks
Unsupervised Multiple-Instance Learning for Instance Search
Unsupervised Representation Learning with Prior-Free and Adversarial Mechanism Embedded Autoencoders
VCF: Velocity Correlation Filter, Towards Space-Borne Satellite Video Tracking
Visual Confusion Label Tree for Image Classification

« prev Page 3/3