Automatic target recognition via sparse representations

Estabridis, Katia

doi:10.1117/12.849591

Cited by 13 publications

(13 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For example, for vehicle recognition, rotation invariance is important and many more vehicle images taken from different angles are crucial. 14 Those are often similar. In such scenarios, there will be more similar SIFT descriptors.…”

Section: Motivationmentioning

confidence: 95%

Clustering-weighted SIFT-based classification method via sparse representation

Sun

2014

J. Electron. Imaging

View full text Add to dashboard Cite

Abstract. In recent years, sparse representation-based classification (SRC) has received significant attention due to its high recognition rate. However, the original SRC method requires a rigid alignment, which is crucial for its application. Therefore, features such as SIFT descriptors are introduced into the SRC method, resulting in an alignment-free method. However, a feature-based dictionary always contains considerable useful information for recognition. We explore the relationship of the similarity of the SIFT descriptors to multitask recognition and propose a clustering-weighted SIFT-based SRC method (CWS-SRC). The proposed approach is considerably more suitable for multitask recognition with sufficient samples. Using two public face databases (AR and Yale face) and a self-built car-model database, the performance of the proposed method is evaluated and compared to that of the SRC, SIFT matching, and MKD-SRC methods. Experimental results indicate that the proposed method exhibits better performance in the alignment-free scenario with sufficient samples. © The Authors. Published by SPIE under a Creative Commons Attribution 3.0 Unported License. Distribution or reproduction of this work in whole or in part requires full attribution of the original publication, including its DOI.

show abstract

Section: Motivationmentioning

confidence: 95%

Clustering-weighted SIFT-based classification method via sparse representation

Sun

2014

J. Electron. Imaging

View full text Add to dashboard Cite

show abstract

“…Principles of Compressed Sensing theory developed by Candes, Romberg and Tao, [4], [5], have been demonstrated to be robust when applied to recognition systems. Wright in [6], formulated the recognition problem as an undetermined system of equations y = Ax , with A containing vectorized image samples from all classes and the solution vector x containing the identity of the test input y. Estabridis, [7], expanded on their work and applied it to target recognition in clutter at ranges of 1, 2 and 3 Kilometers for both EO and IR modalities. Neither work included spatial information during the discrimination process, but both approaches demonstrated good performance within the sparse representation framework.…”

Section: Introductionmentioning

confidence: 99%

“…Neither work included spatial information during the discrimination process, but both approaches demonstrated good performance within the sparse representation framework. Previous work on object identification for faces and ground vehicles based on CS theory ( [6], [8], and [7]), showed the robustness of exploiting sparsity to solve the recognition problem. However, the tested scenarios assume that there is sufficient training data to create robust target representations that extrapolate to future test input data.…”

Section: Introductionmentioning

confidence: 99%

Regional features with adaptable global mappings for recognition systems

Estabridis

2012

2012 Conference Record of the Forty Sixth Asilomar Conference on Signals, Systems and Computers (ASILOMAR)

Self Cite

View full text Add to dashboard Cite

This paper proposes an adaptive recognition system that integrates local and global features while jointly classifying and learning from unlabeled data. Dictionaries based on local descriptors serve as the basis for the recognition system and at the same time provide spatial-mappings derived from the location of the selected features during classification, via l1 minimization techniques. The mappings provide a global object representation that is utilized to discriminate among classes with candidate descriptors. Additionally updating or learning new local descriptors (via non-parametric Bayes) from unlabeled data within a dictionary framework, provides the flexibility needed when training data is limited.KEY WORDS adaptable global mappings, integration of global and regional features, face recognition, unsupervised learning

show abstract

“…One classification algorithm which has received much attention for its success in face recognition is that proposed by Wright et al [1] using methods from the theory of sparse matrix representation and compressive sensing. This method has also been applied successfully to automatic target recognition, specifically to ground vehicles [2]. The sparse representation based classification (SRC) algorithm from [1] relies on representing an image to be classified as a linear combination of elements from an overcomplete dictionary of training images.…”

Section: Introductionmentioning

confidence: 99%

Object recognition in ocean imagery using feature selection and compressive sensing

Rainey

Stastny

2011

2011 IEEE Applied Imagery Pattern Recognition Workshop (AIPR)

View full text Add to dashboard Cite

Ship recognition and classification in electro-optical satellite imagery is a challenging problem with important military applications. The problem is similar to that of face recognition, but with many unique considerations. A ship's appearance can vary dramatically from image to image depending on factors such as lighting condition, sensor angle, and ocean state, and there is often wide variation between ships of the same class. Collecting and labeling sufficient training data is another challenge. We consider how appropriate feature selection and description can assist in addressing these challenges. Our proposed algorithm for vessel classification combines shape invariant features such as SIFT with a well known face recognition algorithm from the theory of sparse representation and compressive sensing. We demonstrate improved classification accuracy using invariant features at significant key points instead of random features to represent images. We also discuss how algorithms such as this are currently implemented to detect and classify ships and other objects in ocean imagery.

show abstract

Automatic target recognition via sparse representations

Cited by 13 publications

References 0 publications

Clustering-weighted SIFT-based classification method via sparse representation

Clustering-weighted SIFT-based classification method via sparse representation

Regional features with adaptable global mappings for recognition systems

Object recognition in ocean imagery using feature selection and compressive sensing

Contact Info

Product

Resources

About