Matches in SemOpenAlex for { <https://semopenalex.org/work/W2800388730> ?p ?o ?g. }
- W2800388730 abstract "It is common to choose image classification network as backbone in the object detector. The art-of-the-state image classification network exhibits excellent performance on image classification, but that network hurts the detection efficiency, mainly due to the coarseness of features from several convolution and pooling layers. In this paper, we present a single deep neural network with inceptions, called StairsNet, to take advantage of the art-of-the-state image classification network in object detection. In contrast to previous single network SSD [13] which uses VGG-16 as a feature to extract network, our approach applies recently state-of-the-art classification network Residual Network (ResNets [5]). Meanwhile, to avoid coarseness of the last CNN feature, StairsNet not only utilizes various of scale features, but also mixes different scale features to predict. To this end, we insert two stairs-like architectures into the network: top stairway network that mixes multi-scale feature maps as input to predict bounding boxes and bottom stairway network that turns into two different scale feature branches. Our StairsNet significantly increases the PASCAL-style mean Average Precision (mAP) from 75.0% (SSD + ResNet-101) to 77.7%. Code is available at https://github.com/gwyve/caffe/tree/StairsNet." @default.
- W2800388730 created "2018-05-17" @default.
- W2800388730 creator A5023847207 @default.
- W2800388730 creator A5032632856 @default.
- W2800388730 creator A5043064822 @default.
- W2800388730 creator A5090204496 @default.
- W2800388730 date "2018-01-01" @default.
- W2800388730 modified "2023-09-28" @default.
- W2800388730 title "StairsNet: Mixed Multi-scale Network for Object Detection" @default.
- W2800388730 cites W2046382188 @default.
- W2800388730 cites W2066624635 @default.
- W2800388730 cites W2068730032 @default.
- W2800388730 cites W2088049833 @default.
- W2800388730 cites W2097117768 @default.
- W2800388730 cites W2100495367 @default.
- W2800388730 cites W2102605133 @default.
- W2800388730 cites W2117539524 @default.
- W2800388730 cites W2168804568 @default.
- W2800388730 cites W2183341477 @default.
- W2800388730 cites W2194775991 @default.
- W2800388730 cites W2565639579 @default.
- W2800388730 cites W2962992847 @default.
- W2800388730 cites W2963037989 @default.
- W2800388730 cites W3106250896 @default.
- W2800388730 cites W7746136 @default.
- W2800388730 doi "https://doi.org/10.1007/978-3-319-77380-3_29" @default.
- W2800388730 hasPublicationYear "2018" @default.
- W2800388730 type Work @default.
- W2800388730 sameAs 2800388730 @default.
- W2800388730 citedByCount "0" @default.
- W2800388730 crossrefType "book-chapter" @default.
- W2800388730 hasAuthorship W2800388730A5023847207 @default.
- W2800388730 hasAuthorship W2800388730A5032632856 @default.
- W2800388730 hasAuthorship W2800388730A5043064822 @default.
- W2800388730 hasAuthorship W2800388730A5090204496 @default.
- W2800388730 hasConcept C115961682 @default.
- W2800388730 hasConcept C138885662 @default.
- W2800388730 hasConcept C153180895 @default.
- W2800388730 hasConcept C154945302 @default.
- W2800388730 hasConcept C193415008 @default.
- W2800388730 hasConcept C199360897 @default.
- W2800388730 hasConcept C2776151529 @default.
- W2800388730 hasConcept C2776401178 @default.
- W2800388730 hasConcept C31258907 @default.
- W2800388730 hasConcept C31972630 @default.
- W2800388730 hasConcept C38652104 @default.
- W2800388730 hasConcept C41008148 @default.
- W2800388730 hasConcept C41895202 @default.
- W2800388730 hasConcept C45347329 @default.
- W2800388730 hasConcept C50644808 @default.
- W2800388730 hasConcept C52622490 @default.
- W2800388730 hasConcept C70437156 @default.
- W2800388730 hasConcept C75294576 @default.
- W2800388730 hasConcept C75608658 @default.
- W2800388730 hasConcept C81363708 @default.
- W2800388730 hasConcept C88796919 @default.
- W2800388730 hasConceptScore W2800388730C115961682 @default.
- W2800388730 hasConceptScore W2800388730C138885662 @default.
- W2800388730 hasConceptScore W2800388730C153180895 @default.
- W2800388730 hasConceptScore W2800388730C154945302 @default.
- W2800388730 hasConceptScore W2800388730C193415008 @default.
- W2800388730 hasConceptScore W2800388730C199360897 @default.
- W2800388730 hasConceptScore W2800388730C2776151529 @default.
- W2800388730 hasConceptScore W2800388730C2776401178 @default.
- W2800388730 hasConceptScore W2800388730C31258907 @default.
- W2800388730 hasConceptScore W2800388730C31972630 @default.
- W2800388730 hasConceptScore W2800388730C38652104 @default.
- W2800388730 hasConceptScore W2800388730C41008148 @default.
- W2800388730 hasConceptScore W2800388730C41895202 @default.
- W2800388730 hasConceptScore W2800388730C45347329 @default.
- W2800388730 hasConceptScore W2800388730C50644808 @default.
- W2800388730 hasConceptScore W2800388730C52622490 @default.
- W2800388730 hasConceptScore W2800388730C70437156 @default.
- W2800388730 hasConceptScore W2800388730C75294576 @default.
- W2800388730 hasConceptScore W2800388730C75608658 @default.
- W2800388730 hasConceptScore W2800388730C81363708 @default.
- W2800388730 hasConceptScore W2800388730C88796919 @default.
- W2800388730 hasLocation W28003887301 @default.
- W2800388730 hasOpenAccess W2800388730 @default.
- W2800388730 hasPrimaryLocation W28003887301 @default.
- W2800388730 hasRelatedWork W2170957789 @default.
- W2800388730 hasRelatedWork W2897317480 @default.
- W2800388730 hasRelatedWork W2937165163 @default.
- W2800388730 hasRelatedWork W3002621688 @default.
- W2800388730 hasRelatedWork W3018964254 @default.
- W2800388730 hasRelatedWork W3086723254 @default.
- W2800388730 hasRelatedWork W3130515555 @default.
- W2800388730 hasRelatedWork W3134487554 @default.
- W2800388730 hasRelatedWork W3142455195 @default.
- W2800388730 hasRelatedWork W3186911986 @default.
- W2800388730 hasRelatedWork W3193706145 @default.
- W2800388730 hasRelatedWork W3196353492 @default.
- W2800388730 hasRelatedWork W3197450834 @default.
- W2800388730 hasRelatedWork W3197801354 @default.
- W2800388730 hasRelatedWork W3199752825 @default.
- W2800388730 hasRelatedWork W3200157107 @default.
- W2800388730 hasRelatedWork W3208438618 @default.
- W2800388730 hasRelatedWork W3208475496 @default.
- W2800388730 hasRelatedWork W3001519082 @default.
- W2800388730 hasRelatedWork W3109992460 @default.