IJCV 2014

107 papers
The International Journal of Computer Vision published 27 papers in 2014, emphasizing spatial reasoning, scene understanding, and automated image annotation. Key research directions included continuous representations for precise spatial orientation and depth-aware reasoning for scene analysis. Notable studies explored fine-grained categorization using hybrid human-machine systems and the application of segmentation propagation for ImageNet auto-annotation. There was also a significant focus on action recognition, specifically regarding the transition from laboratory settings to real-world applications and learning from weakly labeled video data. Additional research addressed domain adaptation for face recognition and the development of hybrid camera systems for high-resolution video acquisition. Prolific authors during this year included Vittorio Ferrari, Serge J. Belongie, and Ashish Kapoor. The year’s publications also covered fundamental tasks such as junction detection in natural images and the discovery of object parts and attributes through relative annotations. These works collectively reflect a trend toward improving robustness in dense environments and bridging the gap between controlled datasets and real-world visual information.

Top Research Directions

Research directions most represented by IJCV 2014 papers.

Modeling Non-Rigid Deformations in Dynamic 3D Scenes 2
Resolving Subspace Intersection Ambiguity in High Dimensional Clustering 1
Modeling Data with Heterogeneous Intrinsic Dimensionality 1
Mapping Geometric Form to Functional Affordance 1
Computational Flat Optics for Ultra Thin Imaging 1
Head Centric Tracking in Occluded Environments 1
Hierarchical Multi-Task Learning for Structured Representation 1
Hierarchical Space Partitioning for Black Box Optimization 1
Recovering Absolute Metric Scale from Human Body Priors 1
Joint Spatio-Temporal Optimization 1
Asymmetric Feature Alignment for Domain Adaptation 1
Extending Depth of Field via Computational Imaging 1
Reconstructing Scenes Outside the Line of Sight 1
Designing for Human AI Complementarity 1
Frequency Domain Diagonalization for Fast Image Restoration 1

Top Authors

AuthorPapers
Ashish Kapoor4
Kristen Grauman4
Serge Belongie4
Alan Yuille3
Andrew Zisserman3
Luc Van Gool3
Nicu Sebe3
Qionghai Dai3
Ramesh Raskar3
Alberto Del Bimbo2
Andrew W Fitzgibbon2
Antonio Torralba2
Brian Kulis2
Carsten Rother2
Chenguang Ma2
Dani Lischinski2
Dong Xu2
Edward Johns2
Eric Horvitz2
Fabio Cuzzolin2

Impactful Papers

Papers from IJCV 2014 that shaped research clusters.

Foundational

Influential

Papers

Title
The Ignorant Led by the Blind: A Hybrid Human-Machine Vision System for Fine-Grained Categorization.
The Shape Boltzmann Machine: A Strong Model of Object Shape
The SUN Attribute Database: Beyond Categories for Deeper Scene Understanding
Ultra-fast Lensless Computational Imaging through 5D Frequency Analysis of Time-resolved Light Transport
Understanding, Optimising, and Extending Data Compression with Anisotropic Diffusion
Visual Saliency with Statistical Priors
Weakly-Supervised Cross-Domain Dictionary Learning for Visual Recognition

« prev Page 3/3