FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3025966429> ?p ?o ?g. }

• W3025966429 endingPage "6758" @default.
• W3025966429 startingPage "6745" @default.
• W3025966429 abstract "Input scale plays an important role in modern detection frameworks, and an optimal training scale for images exists empirically. However, the optimal one usually cannot be reached in facing extremely large images under the memory constraint. In this study, we explore the scale effect inside the object detection pipeline and find that feature upsampling with the introduction of high-resolution information benefits the detection. Compared with direct input upscaling, feature upsampling trades a small performance loss for a large amount of memory savings. From these observations, we propose a self-supervised feature augmentation network, which takes downsampled images as inputs and aims to generate comparable features with the ones when feeding upscaled images to networks. We present a guided feature upsampling module, which takes downsampled images as inputs, to learn upscaled feature representations with the supervision of real large features acquired from upscaled images. In a self-supervised learning manner, we can introduce detailed information of images to the network. For an efficient feature upsampling, we design a residualized sub-pixel convolution block based on a sub-pixel convolution layer, which involves considerable information in upsampling process. Experiments on Mapillary Vistas Dataset (MVD), Cityscapes, and COCO are conducted to demonstrate the effectiveness of our method. On the MVD and Cityscapes detection benchmarks, in which the images are extremely large, our method surpasses current approaches. On COCO, the proposed method obtains comparable results to existing methods but with higher efficiency." @default.
• W3025966429 created "2020-05-21" @default.
• W3025966429 creator A5000680423 @default.
• W3025966429 creator A5021240812 @default.
• W3025966429 creator A5022636178 @default.
• W3025966429 creator A5029314133 @default.
• W3025966429 creator A5036489223 @default.
• W3025966429 creator A5048871276 @default.
• W3025966429 creator A5049927484 @default.
• W3025966429 creator A5067338071 @default.
• W3025966429 creator A5069298091 @default.
• W3025966429 date "2020-01-01" @default.
• W3025966429 modified "2023-10-16" @default.
• W3025966429 title "Self-Supervised Feature Augmentation for Large Image Object Detection" @default.
• W3025966429 cites W1536680647 @default.
• W3025966429 cites W1903029394 @default.
• W3025966429 cites W1932624639 @default.
• W3025966429 cites W2031489346 @default.
• W3025966429 cites W2102605133 @default.
• W3025966429 cites W2117539524 @default.
• W3025966429 cites W219040644 @default.
• W3025966429 cites W2194775991 @default.
• W3025966429 cites W2210803143 @default.
• W3025966429 cites W2216125271 @default.
• W3025966429 cites W2277195237 @default.
• W3025966429 cites W2288122362 @default.
• W3025966429 cites W2340897893 @default.
• W3025966429 cites W2412782625 @default.
• W3025966429 cites W2418346537 @default.
• W3025966429 cites W2476548250 @default.
• W3025966429 cites W2490270993 @default.
• W3025966429 cites W2555182955 @default.
• W3025966429 cites W2558661413 @default.
• W3025966429 cites W2560023338 @default.
• W3025966429 cites W2563705555 @default.
• W3025966429 cites W2565639579 @default.
• W3025966429 cites W2570343428 @default.
• W3025966429 cites W2575671312 @default.
• W3025966429 cites W2592939477 @default.
• W3025966429 cites W2598666589 @default.
• W3025966429 cites W2750549109 @default.
• W3025966429 cites W2753588254 @default.
• W3025966429 cites W2768489488 @default.
• W3025966429 cites W2781228439 @default.
• W3025966429 cites W2798877493 @default.
• W3025966429 cites W2798991696 @default.
• W3025966429 cites W2894651257 @default.
• W3025966429 cites W2947156405 @default.
• W3025966429 cites W2949695917 @default.
• W3025966429 cites W2962824366 @default.
• W3025966429 cites W2962917547 @default.
• W3025966429 cites W2962992847 @default.
• W3025966429 cites W2963016543 @default.
• W3025966429 cites W2963037989 @default.
• W3025966429 cites W2963150697 @default.
• W3025966429 cites W2963420272 @default.
• W3025966429 cites W2963604034 @default.
• W3025966429 cites W2963749571 @default.
• W3025966429 cites W2963813458 @default.
• W3025966429 cites W2963815618 @default.
• W3025966429 cites W2963857746 @default.
• W3025966429 cites W2963881378 @default.
• W3025966429 cites W2964241181 @default.
• W3025966429 cites W2964297960 @default.
• W3025966429 cites W2964330706 @default.
• W3025966429 cites W2964979676 @default.
• W3025966429 cites W2985368596 @default.
• W3025966429 cites W2998655204 @default.
• W3025966429 cites W3098389804 @default.
• W3025966429 cites W343636949 @default.
• W3025966429 doi "https://doi.org/10.1109/tip.2020.2993403" @default.
• W3025966429 hasPublicationYear "2020" @default.
• W3025966429 type Work @default.
• W3025966429 sameAs 3025966429 @default.
• W3025966429 citedByCount "18" @default.
• W3025966429 countsByYear W30259664292019 @default.
• W3025966429 countsByYear W30259664292021 @default.
• W3025966429 countsByYear W30259664292022 @default.
• W3025966429 countsByYear W30259664292023 @default.
• W3025966429 crossrefType "journal-article" @default.
• W3025966429 hasAuthorship W3025966429A5000680423 @default.
• W3025966429 hasAuthorship W3025966429A5021240812 @default.
• W3025966429 hasAuthorship W3025966429A5022636178 @default.
• W3025966429 hasAuthorship W3025966429A5029314133 @default.
• W3025966429 hasAuthorship W3025966429A5036489223 @default.
• W3025966429 hasAuthorship W3025966429A5048871276 @default.
• W3025966429 hasAuthorship W3025966429A5049927484 @default.
• W3025966429 hasAuthorship W3025966429A5067338071 @default.
• W3025966429 hasAuthorship W3025966429A5069298091 @default.
• W3025966429 hasConcept C110384440 @default.
• W3025966429 hasConcept C115961682 @default.
• W3025966429 hasConcept C121332964 @default.
• W3025966429 hasConcept C138885662 @default.
• W3025966429 hasConcept C153180895 @default.
• W3025966429 hasConcept C154945302 @default.
• W3025966429 hasConcept C160633673 @default.
• W3025966429 hasConcept C199360897 @default.
• W3025966429 hasConcept C2524010 @default.

• next