Matches in SemOpenAlex for { <https://semopenalex.org/work/W2105103432> ?p ?o ?g. }
Showing items 1 to 91 of
91
with 100 items per page.
- W2105103432 abstract "Two recent approaches have achieved state-of-the-art results in image captioning. The first uses a pipelined process where a set of candidate words is generated by a convolutional neural network (CNN) trained on images, and then a maximum entropy (ME) language model is used to arrange these words into a coherent sentence. The second uses the penultimate activation layer of the CNN as input to a recurrent neural network (RNN) that then generates the caption sequence. In this paper, we compare the merits of these different language modeling approaches for the first time by using the same state-ofthe-art CNN as input. We examine issues in the different approaches, including linguistic irregularities, caption repetition, and data set overlap. By combining key aspects of the ME and RNN methods, we achieve a new record performance over previously published results on the benchmark COCO dataset. However, the gains we see in BLEU do not translate to human judgments." @default.
- W2105103432 created "2016-06-24" @default.
- W2105103432 creator A5016695723 @default.
- W2105103432 creator A5035017068 @default.
- W2105103432 creator A5040211080 @default.
- W2105103432 creator A5046235098 @default.
- W2105103432 creator A5048159038 @default.
- W2105103432 creator A5057457287 @default.
- W2105103432 creator A5068398083 @default.
- W2105103432 creator A5069954850 @default.
- W2105103432 date "2015-01-01" @default.
- W2105103432 modified "2023-10-18" @default.
- W2105103432 title "Language Models for Image Captioning: The Quirks and What Works" @default.
- W2105103432 cites W1489525520 @default.
- W2105103432 cites W1861492603 @default.
- W2105103432 cites W1895577753 @default.
- W2105103432 cites W1895989618 @default.
- W2105103432 cites W1897761818 @default.
- W2105103432 cites W1905882502 @default.
- W2105103432 cites W1947481528 @default.
- W2105103432 cites W1999965501 @default.
- W2105103432 cites W2101105183 @default.
- W2105103432 cites W2117539524 @default.
- W2105103432 cites W2133459682 @default.
- W2105103432 cites W2146574666 @default.
- W2105103432 cites W2157331557 @default.
- W2105103432 cites W2171361956 @default.
- W2105103432 cites W2250489405 @default.
- W2105103432 cites W2951183276 @default.
- W2105103432 cites W2962706528 @default.
- W2105103432 cites W2962835968 @default.
- W2105103432 cites W68733909 @default.
- W2105103432 doi "https://doi.org/10.3115/v1/p15-2017" @default.
- W2105103432 hasPublicationYear "2015" @default.
- W2105103432 type Work @default.
- W2105103432 sameAs 2105103432 @default.
- W2105103432 citedByCount "137" @default.
- W2105103432 countsByYear W21051034322015 @default.
- W2105103432 countsByYear W21051034322016 @default.
- W2105103432 countsByYear W21051034322017 @default.
- W2105103432 countsByYear W21051034322018 @default.
- W2105103432 countsByYear W21051034322019 @default.
- W2105103432 countsByYear W21051034322020 @default.
- W2105103432 countsByYear W21051034322021 @default.
- W2105103432 countsByYear W21051034322022 @default.
- W2105103432 countsByYear W21051034322023 @default.
- W2105103432 crossrefType "proceedings-article" @default.
- W2105103432 hasAuthorship W2105103432A5016695723 @default.
- W2105103432 hasAuthorship W2105103432A5035017068 @default.
- W2105103432 hasAuthorship W2105103432A5040211080 @default.
- W2105103432 hasAuthorship W2105103432A5046235098 @default.
- W2105103432 hasAuthorship W2105103432A5048159038 @default.
- W2105103432 hasAuthorship W2105103432A5057457287 @default.
- W2105103432 hasAuthorship W2105103432A5068398083 @default.
- W2105103432 hasAuthorship W2105103432A5069954850 @default.
- W2105103432 hasBestOaLocation W21051034321 @default.
- W2105103432 hasConcept C115961682 @default.
- W2105103432 hasConcept C138885662 @default.
- W2105103432 hasConcept C154945302 @default.
- W2105103432 hasConcept C157657479 @default.
- W2105103432 hasConcept C204321447 @default.
- W2105103432 hasConcept C28490314 @default.
- W2105103432 hasConcept C41008148 @default.
- W2105103432 hasConcept C41895202 @default.
- W2105103432 hasConceptScore W2105103432C115961682 @default.
- W2105103432 hasConceptScore W2105103432C138885662 @default.
- W2105103432 hasConceptScore W2105103432C154945302 @default.
- W2105103432 hasConceptScore W2105103432C157657479 @default.
- W2105103432 hasConceptScore W2105103432C204321447 @default.
- W2105103432 hasConceptScore W2105103432C28490314 @default.
- W2105103432 hasConceptScore W2105103432C41008148 @default.
- W2105103432 hasConceptScore W2105103432C41895202 @default.
- W2105103432 hasLocation W21051034321 @default.
- W2105103432 hasLocation W21051034322 @default.
- W2105103432 hasLocation W21051034323 @default.
- W2105103432 hasOpenAccess W2105103432 @default.
- W2105103432 hasPrimaryLocation W21051034321 @default.
- W2105103432 hasRelatedWork W2503073734 @default.
- W2105103432 hasRelatedWork W2547835662 @default.
- W2105103432 hasRelatedWork W2596543464 @default.
- W2105103432 hasRelatedWork W2891852518 @default.
- W2105103432 hasRelatedWork W2905654560 @default.
- W2105103432 hasRelatedWork W2923366293 @default.
- W2105103432 hasRelatedWork W3008515501 @default.
- W2105103432 hasRelatedWork W3183824823 @default.
- W2105103432 hasRelatedWork W4320016117 @default.
- W2105103432 hasRelatedWork W2519434724 @default.
- W2105103432 isParatext "false" @default.
- W2105103432 isRetracted "false" @default.
- W2105103432 magId "2105103432" @default.
- W2105103432 workType "article" @default.