FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2938544580> ?p ?o ?g. }

• W2938544580 abstract "Multi-task learning is commonly used in autonomous driving for solving various visual perception tasks. It offers significant benefits in terms of both performance and computational complexity. Current work on multi-task learning networks focus on processing a single input image and there is no known implementation of multi-task learning handling a sequence of images. In this work, we propose a multi-stream multi-task network to take advantage of using feature representations from preceding frames in a video sequence for joint learning of segmentation, depth, and motion. The weights of the current and previous encoder are shared so that features computed in the previous frame can be leveraged without additional computation. In addition, we propose to use the geometric mean of task losses as a better alternative to the weighted average of task losses. The proposed loss function facilitates better handling of the difference in convergence rates of different tasks. Experimental results on KITTI, Cityscapes and SYNTHIA datasets demonstrate that the proposed strategies outperform various existing multi-task learning solutions." @default.
• W2938544580 created "2019-04-25" @default.
• W2938544580 creator A5014764449 @default.
• W2938544580 creator A5020528946 @default.
• W2938544580 creator A5074948543 @default.
• W2938544580 creator A5091875799 @default.
• W2938544580 date "2019-04-15" @default.
• W2938544580 modified "2023-09-23" @default.
• W2938544580 title "MultiNet++: Multi-Stream Feature Aggregation and Geometric Loss Strategy for Multi-Task Learning." @default.
• W2938544580 cites W1485009520 @default.
• W2938544580 cites W1499332833 @default.
• W2938544580 cites W1522301498 @default.
• W2938544580 cites W1738019091 @default.
• W2938544580 cites W1903029394 @default.
• W2938544580 cites W1923404803 @default.
• W2938544580 cites W1983364832 @default.
• W2938544580 cites W2016053056 @default.
• W2938544580 cites W2060727197 @default.
• W2938544580 cites W2064675550 @default.
• W2938544580 cites W2115579991 @default.
• W2938544580 cites W2115761837 @default.
• W2938544580 cites W2156303437 @default.
• W2938544580 cites W2194775991 @default.
• W2938544580 cites W2251743902 @default.
• W2938544580 cites W2286655030 @default.
• W2938544580 cites W2292439652 @default.
• W2938544580 cites W2308045930 @default.
• W2938544580 cites W2340897893 @default.
• W2938544580 cites W2431874326 @default.
• W2938544580 cites W2508429489 @default.
• W2938544580 cites W2514012460 @default.
• W2938544580 cites W2548527721 @default.
• W2938544580 cites W2557283755 @default.
• W2938544580 cites W2562137921 @default.
• W2938544580 cites W2581955877 @default.
• W2938544580 cites W2604455318 @default.
• W2938544580 cites W2609930246 @default.
• W2938544580 cites W2612131472 @default.
• W2938544580 cites W2618011341 @default.
• W2938544580 cites W2734515856 @default.
• W2938544580 cites W2770902190 @default.
• W2938544580 cites W2774058387 @default.
• W2938544580 cites W2784231336 @default.
• W2938544580 cites W2789102901 @default.
• W2938544580 cites W2793939440 @default.
• W2938544580 cites W2890538051 @default.
• W2938544580 cites W2895340898 @default.
• W2938544580 cites W2895387432 @default.
• W2938544580 cites W2904631343 @default.
• W2938544580 cites W2910695916 @default.
• W2938544580 cites W2911354217 @default.
• W2938544580 cites W2913340405 @default.
• W2938544580 cites W2915692735 @default.
• W2938544580 cites W2950307714 @default.
• W2938544580 cites W2962743139 @default.
• W2938544580 cites W2962945654 @default.
• W2938544580 cites W2963168538 @default.
• W2938544580 cites W2963263064 @default.
• W2938544580 cites W2963323244 @default.
• W2938544580 cites W2963377935 @default.
• W2938544580 cites W2963498646 @default.
• W2938544580 cites W2963677766 @default.
• W2938544580 cites W2963833733 @default.
• W2938544580 cites W2964121793 @default.
• W2938544580 cites W2964191259 @default.
• W2938544580 cites W2964262254 @default.
• W2938544580 cites W2964286567 @default.
• W2938544580 cites W3105202226 @default.
• W2938544580 cites W3106338485 @default.
• W2938544580 cites W2795042520 @default.
• W2938544580 hasPublicationYear "2019" @default.
• W2938544580 type Work @default.
• W2938544580 sameAs 2938544580 @default.
• W2938544580 citedByCount "7" @default.
• W2938544580 countsByYear W29385445802019 @default.
• W2938544580 countsByYear W29385445802020 @default.
• W2938544580 countsByYear W29385445802021 @default.
• W2938544580 crossrefType "posted-content" @default.
• W2938544580 hasAuthorship W2938544580A5014764449 @default.
• W2938544580 hasAuthorship W2938544580A5020528946 @default.
• W2938544580 hasAuthorship W2938544580A5074948543 @default.
• W2938544580 hasAuthorship W2938544580A5091875799 @default.
• W2938544580 hasConcept C111919701 @default.
• W2938544580 hasConcept C118505674 @default.
• W2938544580 hasConcept C119857082 @default.
• W2938544580 hasConcept C138885662 @default.
• W2938544580 hasConcept C14036430 @default.
• W2938544580 hasConcept C153180895 @default.
• W2938544580 hasConcept C154945302 @default.
• W2938544580 hasConcept C162324750 @default.
• W2938544580 hasConcept C169760540 @default.
• W2938544580 hasConcept C187736073 @default.
• W2938544580 hasConcept C26760741 @default.
• W2938544580 hasConcept C2776401178 @default.
• W2938544580 hasConcept C2777303404 @default.
• W2938544580 hasConcept C2778112365 @default.
• W2938544580 hasConcept C2780451532 @default.
• W2938544580 hasConcept C28006648 @default.
• W2938544580 hasConcept C41008148 @default.
• W2938544580 hasConcept C41895202 @default.

• next