A distributed SVM for image annotation

Alham, Nasullah Khalid; Li, Maozhen; Hammoud, Suhel; Liu, Yang; Ponraj, Mahesh

doi:10.1109/fskd.2010.5569084

Cited by 5 publications

(2 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Figure 4 shows the efficiency of the MRMSVM in SVM training which achieved close to 12 times in speedup. Figure 5 shows the efficiency of the MRMSVM in comparison with MRSMO [27] which is one against all based distributed multiclass SVM. MRMSVM is more efficient due to the fact that training data for each binary classifier is a subset of the available training data which only contains the data for the two classes involved.…”

Section: B Performance Evaluationmentioning

confidence: 99%

Parallelizing multiclass Support Vector Machines for scalable image annotation

Alham

Liu

et al. 2011

2011 Eighth International Conference on Fuzzy Systems and Knowledge Discovery (FSKD)

Self Cite

View full text Add to dashboard Cite

Machine learning techniques have facilitated image retrieval by automatically classifying and annotating images with keywords. Among them Support Vector Machines (SVMs) are used extensively due to their generalization properties. SVM was initially designed for binary classifications. However, most classification problems arising in domains such as image annotation usually involve more than two classes. Notably SVM training is a computationally intensive process especially when the training dataset is large. This paper presents MRMSVM, a distributed multiclass SVM algorithm for large scale image annotation which partitions the training dataset into smaller binary chunks and train SVM in parallel using a cluster of computers. MRMSVM is evaluated in an experimental environment showing that the distributed Multiclass SVM algorithm reduces the training time significantly while maintaining a high level of accuracy in classifications.

show abstract

Section: B Performance Evaluationmentioning

confidence: 99%

Parallelizing multiclass Support Vector Machines for scalable image annotation

Alham

Liu

et al. 2011

2011 Eighth International Conference on Fuzzy Systems and Knowledge Discovery (FSKD)

Self Cite

View full text Add to dashboard Cite

show abstract

“…Catanzaro et al [41] proposed a parallel SMO algorithm based on MapReduce, but this algorithm is somewhat inefficient because of the iterative nature of SMO. In [42], Alham discussed an efficient and scalable approach based on a single MapReduce phase. They announced that this method has minimal data movement between nodes and that it also minimizes communication overheads.…”

Section: Image Classificationmentioning

confidence: 99%