ICME 2014

135 papers
The 2014 IEEE International Conference on Multimedia and Expo (ICME) featured 185 papers focusing on human-centric analysis and the integration of visual context. Key research directions included skeleton-based action recognition, articulated human detection, and the use of temporal context to resolve visual ambiguity. Methodological trends highlighted the application of graph-based manifold ranking for visual tracking, hierarchical L0 sparsity for image decomposition, and low-rank matrix completion for restoring corrupted motion capture data. In addition to computer vision, the conference addressed multimedia networking and data analytics. Notable work included research on DASH streaming in multirate wireless networks and interactive visual analytics for surveillance event detection. Other studies explored image classification via adaptive pooling and graph-density-based vocabularies for image retrieval. Prolific authors contributing to the year's program included Wen Gao, Wenwu Zhu, Xuelong Li, and Tao Mei. The proceedings reflect a balanced focus on foundational algorithms and practical applications in surveillance, communication, and image processing.

Top Research Directions

Research directions most represented by ICME 2014 papers.

Variance Partitioning for Representational Analysis 1
The Inherent Noise Robustness of Pretrained Models 1
Physical Continuity Priors for Motion and Blur Estimation 1
Relative Spatial Encoding for Coordinate Invariant Representations 1
Overcoming the Staircase Effect in Discrete Depth Representations 1
The Information Loss of Visual Feature Quantization 1
Total Variation Regularization for Edge Preserving Reconstruction 1
Resolving Kinematic Ambiguity in Skeleton Based Action Recognition 1
Spatially Adaptive Regularization 1
Saturation Aware Image Restoration 1
Integrating Physical Degradation Models into Deep Image Restoration 1
Upstream Detection Bottlenecks in Structured Visual Reasoning 1
The Bag of Visual Words Representation 1
Resolving Static Pose Ambiguity with Temporal Context 1
Reconstructing Complete Geometry from Partial Point Clouds 1

Top Authors

AuthorPapers
Wei Feng4
Risheng Liu3
Siwei Ma3
Tao Mei3
Wenjun Zeng3
Wenwu Zhu3
Xuelong Li3
Cha Zhang2
Chi-Man Pun2
Dingwen Zhang2
Guangtao Zhai2
Haipeng Dai2
Jean-Charles Bazin2
Jia Jia2
Jiawan Zhang2
Junwei Han2
Lei Guo2
Liang Wan2
LU FANG2
Mingli Song2

Impactful Papers

Papers from ICME 2014 that shaped research clusters.

Foundational

Influential

Papers

Title
Quality prediction of asymmetrically distorted stereoscopic images from single views
Query difficulty estimation via pseudo relevance feedback for image search
Rank learning on training set selection and image quality assessment
Real-time control of 3D facial animation
Realtime gaze estimation with online calibration
Recommendation on Flickr by combining community user ratings and item importance
Registration of multiple RGBD cameras via local rigid transformations
Representative local features mining for large-scale near-duplicates retrieval
Restoring corrupted motion capture data via jointly low-rank matrix completion
RiskWheel: Interactive visual analytics for surveillance event detection
Robust visual tracking using latent subspace projection pursuit
Saliency detection based on feature learning using Deep Boltzmann Machines
Saliency preserving decolorization
Salient region detection : Integrate both global and local cues
Scalable forest hashing for fast similarity search
Scene text recognition in multiple frames based on text tracking
Screen content coding using non-square intra block copy for HEVC
Secure media sharing in the cloud: Two-dimensional-scalable access control and comprehensive key management
Single depth image super resolution and denoising via coupled dictionary learning with local constraints and shock filtering
Spectral salient object detection
Speeding uplow rank matrix recovery for foreground separation in surveillance videos
Statistical background subtraction based on imbalanced learning
Sub-sampled dictionaries for coarse-to-fine sparse representation-based human action recognition
Supervised trace lasso for robust face recognition
Tag completion with defective tag assignments via image-tag re-weighting
Towards time-varying music auto-tagging based on CAL500 expansion
User specific friend recommendation in social media community
Using mid-high level cues to detect salient object
Video face beautification
Video Face Clustering via Constrained Sparse Representation
Video Summarization based on Nonnegative Linear Reconstruction
Visual attention computation in video of driving environment
Visual dictionaries in the Brain: Comparing HMAX and BOW
Visual tracking via graph-based efficient manifold ranking with low-dimensional compressive features
Who missed the class? - Unifying multi-face detection, tracking and recognition in videos

« prev Page 3/3