CONF grangier:2006:icann/IDIAP A Neural Network to Retrieve Images from Text Queries Grangier, David Bengio, Samy EXTERNAL https://publications.idiap.ch/attachments/reports/2006/grangier_icann06.pdf PUBLIC https://publications.idiap.ch/index.php/publications/showcite/grangier:2006:idiap-06-33 Related documents International Conference on Artificial Neural Networks (ICANN) 2 24-34 2006 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. REPORT grangier:2006:idiap-06-33/IDIAP A Neural Network to Retrieve Images from Text Queries Grangier, David Bengio, Samy EXTERNAL https://publications.idiap.ch/attachments/reports/2006/grangier_rr06-33.pdf PUBLIC Idiap-RR-33-2006 2006 IDIAP 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.