Insight Centre for Data Analytics (DCU) at TRECVid 2014: Instance Search and Semantic Indexing Tasks
Year: 2014
Type: other
Abstract: Insight-DCU participated in the instance search (INS) and semantic indexing (SIN) tasks in 2014.\nTwo very different approaches were submitted for instance search, one based on features extracted using\npre-trained deep convolutional neural networks (CNNs), and another based on local SIFT features, large\nvocabulary visual bag-of-words aggregation, inverted index-based lookup, and geometric verification\non the top-N retrieved results. Two interactive runs and two automatic runs were submitted, the best\ninteractive runs achieved a mAP of 0.135 and the best automatic 0.12. Our semantic indexing runs were\nbased also on using convolutional neural network features, and on Support Vector Machine classifiers\nwith linear and RBF kernels. One run was submitted to the main task, two to the no annotation task,\nand one to the progress task. Data for the no-annotation task was gathered from Google Images and\nImageNet. The main task run has achieved a mAP of 0.086, the best no-annotation runs had a close\nperformance to the main run by achieving a mAP of 0.080, while the progress run had 0.043.
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
Funder: Science Foundation Ireland
OpenAlex ID: W1555695694