FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3136503179> ?p ?o ?g. }

• W3136503179 abstract "This work explores how to design a single neural network capable of adapting to multiple heterogeneous vision tasks, such as image segmentation, 3D detection, and video recognition. This goal is challenging because both network architecture search (NAS) spaces and methods in different tasks are inconsistent. We solve this challenge from both sides. We first introduce a unified design space for multiple tasks and build a multitask NAS benchmark (NAS-Bench-MR) on many widely used datasets, including ImageNet, Cityscapes, KITTI, and HMDB51. We further propose Network Coding Propagation (NCP), which back-propagates gradients of neural predictors to directly update architecture codes along the desired gradient directions to solve various tasks. In this way, optimal architecture configurations can be found by NCP in our large search space in seconds. Unlike prior arts of NAS that typically focus on a single task, NCP has several unique benefits. (1) NCP transforms architecture optimization from data-driven to architecture-driven, enabling joint search an architecture among multitasks with different data distributions. (2) NCP learns from network codes but not original data, enabling it to update the architecture efficiently across datasets. (3) In addition to our NAS-Bench-MR, NCP performs well on other NAS benchmarks, such as NAS-Bench-201. (4) Thorough studies of NCP on inter-, cross-, and intra-tasks highlight the importance of cross-task neural architecture design, i.e., multitask neural architectures and architecture transferring between different tasks. Code is available at https://github.com/dingmyu/NCP." @default.
• W3136503179 created "2021-03-29" @default.
• W3136503179 creator A5022382771 @default.
• W3136503179 creator A5030033393 @default.
• W3136503179 creator A5031223720 @default.
• W3136503179 creator A5042544148 @default.
• W3136503179 creator A5056053058 @default.
• W3136503179 creator A5075158591 @default.
• W3136503179 creator A5075880303 @default.
• W3136503179 creator A5085349794 @default.
• W3136503179 date "2021-03-24" @default.
• W3136503179 modified "2023-09-23" @default.
• W3136503179 title "Learning Versatile Neural Architectures by Propagating Network Codes" @default.
• W3136503179 cites W2126579184 @default.
• W3136503179 cites W2150066425 @default.
• W3136503179 cites W2165681080 @default.
• W3136503179 cites W2184393491 @default.
• W3136503179 cites W2194775991 @default.
• W3136503179 cites W2412782625 @default.
• W3136503179 cites W2560023338 @default.
• W3136503179 cites W2737258237 @default.
• W3136503179 cites W2810075754 @default.
• W3136503179 cites W2905672847 @default.
• W3136503179 cites W2916118939 @default.
• W3136503179 cites W2925303509 @default.
• W3136503179 cites W2952542841 @default.
• W3136503179 cites W2954234207 @default.
• W3136503179 cites W2955425717 @default.
• W3136503179 cites W2958458355 @default.
• W3136503179 cites W2962746461 @default.
• W3136503179 cites W2962861284 @default.
• W3136503179 cites W2962919941 @default.
• W3136503179 cites W2963136578 @default.
• W3136503179 cites W2963163009 @default.
• W3136503179 cites W2963216850 @default.
• W3136503179 cites W2964259004 @default.
• W3136503179 cites W2964294659 @default.
• W3136503179 cites W2964444661 @default.
• W3136503179 cites W2966182616 @default.
• W3136503179 cites W2967733054 @default.
• W3136503179 cites W2968296999 @default.
• W3136503179 cites W2971137988 @default.
• W3136503179 cites W2972830778 @default.
• W3136503179 cites W2980137827 @default.
• W3136503179 cites W2981701755 @default.
• W3136503179 cites W2994749257 @default.
• W3136503179 cites W2994985593 @default.
• W3136503179 cites W2995727387 @default.
• W3136503179 cites W2998003319 @default.
• W3136503179 cites W2999270366 @default.
• W3136503179 cites W3012995654 @default.
• W3136503179 cites W3014641072 @default.
• W3136503179 cites W3030728803 @default.
• W3136503179 cites W3034429256 @default.
• W3136503179 cites W3034704453 @default.
• W3136503179 cites W3034959114 @default.
• W3136503179 cites W3035087309 @default.
• W3136503179 cites W3041574813 @default.
• W3136503179 cites W3043093753 @default.
• W3136503179 cites W3049708585 @default.
• W3136503179 cites W3081305497 @default.
• W3136503179 cites W3094801149 @default.
• W3136503179 cites W3096533519 @default.
• W3136503179 cites W3099431291 @default.
• W3136503179 cites W3107106605 @default.
• W3136503179 cites W3109946440 @default.
• W3136503179 cites W3118608800 @default.
• W3136503179 cites W3120629159 @default.
• W3136503179 doi "https://doi.org/10.48550/arxiv.2103.13253" @default.
• W3136503179 hasPublicationYear "2021" @default.
• W3136503179 type Work @default.
• W3136503179 sameAs 3136503179 @default.
• W3136503179 citedByCount "1" @default.
• W3136503179 countsByYear W31365031792021 @default.
• W3136503179 crossrefType "posted-content" @default.
• W3136503179 hasAuthorship W3136503179A5022382771 @default.
• W3136503179 hasAuthorship W3136503179A5030033393 @default.
• W3136503179 hasAuthorship W3136503179A5031223720 @default.
• W3136503179 hasAuthorship W3136503179A5042544148 @default.
• W3136503179 hasAuthorship W3136503179A5056053058 @default.
• W3136503179 hasAuthorship W3136503179A5075158591 @default.
• W3136503179 hasAuthorship W3136503179A5075880303 @default.
• W3136503179 hasAuthorship W3136503179A5085349794 @default.
• W3136503179 hasBestOaLocation W31365031791 @default.
• W3136503179 hasConcept C105795698 @default.
• W3136503179 hasConcept C108583219 @default.
• W3136503179 hasConcept C113775141 @default.
• W3136503179 hasConcept C118524514 @default.
• W3136503179 hasConcept C119857082 @default.
• W3136503179 hasConcept C123657996 @default.
• W3136503179 hasConcept C13280743 @default.
• W3136503179 hasConcept C142362112 @default.
• W3136503179 hasConcept C153349607 @default.
• W3136503179 hasConcept C154945302 @default.
• W3136503179 hasConcept C162324750 @default.
• W3136503179 hasConcept C177264268 @default.
• W3136503179 hasConcept C179518139 @default.
• W3136503179 hasConcept C185798385 @default.
• W3136503179 hasConcept C187736073 @default.
• W3136503179 hasConcept C193415008 @default.

• next