Abstract | ||
---|---|---|
This work presents a neural network for the retrieval of images from text queries. The proposed network is composed of two main modules: the first one extracts a global picture representation from local block descriptors while the second one aims at solving the retrieval problem from the extracted representation. Both modules are trained jointly to minimize a loss related to the retrieval performance. This approach is shown to be advantageous when compared to previous models relying on unsupervised feature extraction: average precision over Corel queries reaches 26.2% for our model, which should be compared to 21.6% for PAMIR, the best alternative. |
Year | DOI | Venue |
---|---|---|
2006 | 10.1007/11840930_3 | ICANN (2) |
Keywords | Field | DocType |
proposed network,neural network,corel query,retrieval performance,global picture representation,retrieval problem,local block descriptors,best alternative,text query,main module,average precision,feature extraction,machine vision | Data mining,Database query,Pattern recognition,Computer science,Local binary patterns,Image processing,Image retrieval,Feature extraction,Artificial intelligence,Probabilistic latent semantic analysis,Artificial neural network | Conference |
Volume | ISSN | ISBN |
4132 | 0302-9743 | 3-540-38871-0 |
Citations | PageRank | References |
6 | 0.59 | 14 |
Authors | ||
2 |
Name | Order | Citations | PageRank |
---|---|---|---|
David Grangier | 1 | 816 | 41.60 |
Samy Bengio | 2 | 7213 | 485.82 |