近两年顶级会议上关于Distance Metric Learning的paper清单
上月这位学长从外地回到上海，在与导师及其同学共进晚餐时提起了他攻读研究生期间选择的专业领域是DML。他仍然认为这一领域目前研究难度较大，在过去几年间始终未有重大的突破，并基于直觉判断目前关于DML领域的最新论文数量有限。导师若有所思地未作回应。
回头整理了一些相关知识，
梳理了过去两年顶级会议中关于度量学习的论文，
数量庞大，
引起了广泛关注。
其中看上去是不计成本地收集新概念的文章，
似乎近年来DML的发展趋势与早年稀疏编码有几分相似。
Information-based semi-supervised metric learning via entropy-based regularization
一种混合型算法用于解决凸半定优化问题
Similarity Learning for Provably Accurate Sparse Linear Classification
Developing discriminant Fisher kernels
Learning multi-view neighborhood preserving projections
Order Estimation and L0/L1 Regularized Metric Learning for Adaptive Visual Tracking
Non- sparse linear representations are utilized for visual tracking by incorporating an online reservoir metric learning mechanism.
Self-supervised Metric Integration through Cross Diffusion
Learning Hierarchical Similarity Metrics
Large Scale Metric Learning from Equivalence Constraints
Neighborhood Repulsed Metric Learning for Kinship Verification
Developing a robust and discriminative multi-instance distance metric for efficient video categorization.
PCCA: innovative methodology for distance-based learning from pairwise sparse constraints
Collective Motion Resultant Metrics for Averaging and Clustering Techniques of Linear Dynamical Systems with Applications in Analyzing Dynamic Visual Scenarios
A Scalable Dual Approach to Semidefinite Metric Learning
Adaboost在低秩半正定矩阵上的应用及其用于度量学习及计算机辅助诊断研究
Adaptive Metric Differential Tracking (HUST)
Tracking Low Resolution Objects by Metric Preservation (HUST)
Optimal Semi-Supervised Metric Learning for Image Retrieval
Low Rank Metric Learning for Social Image Retrieval
An activity-driven person identification approach through sparse coding technique and discriminant metric learning approach.
This paper presents a deep non-linear metric learning approach combined with independent subspace analysis, specifically applied to face verification tasks.
ACM MM 2011
Biased Metric Learning for Person-Independent Head Pose Estimation
Machine Learning ensembles of sparse distance metrics are employed to perform classification tasks and achieve dimensionality reduction.
Unsupervised Metric Learning for Face Identification in TV Video
Random Ensemble Metrics for Object Recognition
Developing non-linear distance metrics through neural networks in regression tasks, with application to robust human age estimation.
Develop parameterized histogram-based kernel functions on the simplex manifold for image and action recognition.
Metric-Based Learning Approaches for Massive Image Classification Tasks: Generalizing to Unseen Categories with Minimal Computational Overhead
Dual-force Metric Learning for Robust Distractor Resistant Tracker
For the purpose of matching visual traits, learning should be conducted in the metric space of covariance matrices.
Image Annotation Using Metric Learning in Semantic Neighbourhoods
Measuring Image Distances via Embedding in a Semantic Manifold
These Supervised EMD Models are effectively employed in various Supervised Learning Tasks, particularly within the domain of Computer Vision.
Developing a method to establish image distance between classes through large margin and L1-norm regularization.
Image Labeling by Incorporating Sparse Multi-Modal Distance Learning and Contextual Semantic Modeling
Distance Metric Learning Under Covariate Shift
Learning a Distance Metric by Empirical Loss Minimization
Efficiently learning a distance metric for large margin nearest neighbor classification.
Learning a Distance Metric from a Network
Constructing a Hierarchical Structure of Measurements with Distinct Visual Features
Metric Learning with Multiple Kernels
Random Forests for Metric Learning with Implicit Pairwise Position Dependence
**WSDM 2011
**
Extracting User-Generated Media using Distance Metric Learning Approach to Automate Image Tagging