Abstract:Content-based human motion retrieval (CBMR) has been more and more important with the rapid growth of motion capture data, but the gap between high-level semantic concepts and low-level features hinders further performance improvement. Relevance feedback is an effective tool to narrow the semantic gap and enhance the retrieval performance. However, as a type of variable-length multivariate time series (VLMTS), motion capture data has its own characteristics including high-dimensionality, demand for elastic matching, and difficulty representing different movements in a uniform feature space, which make it much more challenging to design an effective relevance feedback approach. This paper presents a novel boosting approach for CBMR and the main contributions include three aspects. First, to fit in with the characteristics of VLMTS data and meet the real-time requirement of relevance feedback, the ensemble learning framework RankBoost is introduced and k-nearest neighbors combining with dynamic time warping (KNN-DTW) is employed as its weak ranker. Second, the set of extended Boolean geometry features containing much richer geometry elements and measures is used to represent motion content, and it provides a comparatively complete feature set for designing the weak ranker of RankBoost. Third, to solve the over-fitting problem caused by the small-sample training of relevance feedback, a novel learning objective composed of minimizing empirical ranking loss and minimizing the maximum generalization loss is proposed for RankBoost ensemble learning. Experimental results on CMU database and its extended database verify the effectiveness of the proposed approach.

A Relevance Feedback Algorith m for Motion Data Retrieval

Relevance Feedback for Human Motion Retrieval Using a Boosting Approach.

A Boosting Method Based on SVM for Relevance Feedback in Content-Based 3D Model Retrieval

A Survey on Relevance Feedback Techniques in Content-based Medical Image Retrieval

A Novel Music Retrieval System with Relevance Feedback

Relevance Feature Mapping for Content-Based Multimedia Information Retrieval

Relevance Feedback In An Adaptive Space With One-Class Svm For Content Based Music Retrieval

Semi-supervised SVM Batch Mode Active Learning for Image Retrieval

Relevance Feedback in Content-Based Image Retrieval: Bayesian Framework, Feature Subspaces, and Progressive Learning

Asymmetric Bagging and Random Subspace for Support Vector Machines-Based Relevance Feedback in Image Retrieval

Evaluation of relevance feedback methods for 3D model retrieval

A Powerful Relevance Feedback Mechanism for Content-Based 3D Model Retrieval

Relevance Feedback: a Power Tool for Interactive Content-Based Image Retrieval

A multimedia retrieval framework based on semi-supervised ranking and relevance feedback.

Video Retrieval Relevance Feedback Algorithm Based on Bayesian Learning

Face Retrieval with Relevance Feedback Using Lifting Wavelets Features

Motion Region-Based Trajectory Analysis and Re-Ranking for Video Retrieval

Relevance Feedback Techniques In Interactive Content-Based Image Retrieval

An Improved Relevance Feedback in CBIR

Relevance feedback for real-world human action retrieval

Weighted distance approach to relevance feedback