%0 Generic %1 li2016learning %A Li, Ang %A Jabri, Allan %A Joulin, Armand %A van der Maaten, Laurens %D 2016 %K %T Learning Visual N-Grams from Web Data %U http://arxiv.org/abs/1612.09161 %X Real-world image recognition systems need to recognize tens of thousands of classes that constitute a plethora of visual concepts. The traditional approach of annotating thousands of images per class for training is infeasible in such a scenario, prompting the use of webly supervised data. This paper explores the training of image-recognition systems on large numbers of images and associated user comments. In particular, we develop visual n-gram models that can predict arbitrary phrases that are relevant to the content of an image. Our visual n-gram models are feed-forward convolutional networks trained using new loss functions that are inspired by n-gram models commonly used in language modeling. We demonstrate the merits of our models in phrase prediction, phrase-based image retrieval, relating images and captions, and zero-shot transfer.

@misc{li2016learning, abstract = {Real-world image recognition systems need to recognize tens of thousands of classes that constitute a plethora of visual concepts. The traditional approach of annotating thousands of images per class for training is infeasible in such a scenario, prompting the use of webly supervised data. This paper explores the training of image-recognition systems on large numbers of images and associated user comments. In particular, we develop visual n-gram models that can predict arbitrary phrases that are relevant to the content of an image. Our visual n-gram models are feed-forward convolutional networks trained using new loss functions that are inspired by n-gram models commonly used in language modeling. We demonstrate the merits of our models in phrase prediction, phrase-based image retrieval, relating images and captions, and zero-shot transfer.}, added-at = {2017-04-14T01:49:28.000+0200}, author = {Li, Ang and Jabri, Allan and Joulin, Armand and van der Maaten, Laurens}, biburl = {https://www.bibsonomy.org/bibtex/253e9cc1fb38a82dd68f6cfad2548282d/ang}, interhash = {64ff94c58e4248185ee7236cc521be0a}, intrahash = {53e9cc1fb38a82dd68f6cfad2548282d}, keywords = {}, note = {cite arxiv:1612.09161}, timestamp = {2017-04-14T01:49:28.000+0200}, title = {Learning Visual N-Grams from Web Data}, url = {http://arxiv.org/abs/1612.09161}, year = 2016 }