FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3010420410> ?p ?o ?g. }

• W3010420410 abstract "Semantic segmentation has achieved remarkable progress but remains challenging due to the complex scene, object occlusion, and so on. Some research works have attempted to use extra information such as a depth map to help RGB based semantic segmentation because the depth map could provide complementary geometric cues. However, due to the inaccessibility of depth sensors, depth information is usually unavailable for the test images. In this paper, we leverage only the depth of training images as the privileged information to mine the hard pixels in semantic segmentation, in which depth information is only available for training images but not available for test images. Specifically, we propose a novel Loss Weight Module, which outputs a loss weight map by employing two depth-related measurements of hard pixels: Depth Prediction Error and Depthaware Segmentation Error. The loss weight map is then applied to segmentation loss, with the goal of learning a more robust model by paying more attention to the hard pixels. Besides, we also explore a curriculum learning strategy based on the loss weight map. Meanwhile, to fully mine the hard pixels on different scales, we apply our loss weight module to multi-scale side outputs. Our hard pixels mining method achieves the state-of-the-art results on two benchmark datasets, and even outperforms the methods which need depth input during testing." @default.
• W3010420410 created "2020-03-13" @default.
• W3010420410 creator A5032618817 @default.
• W3010420410 creator A5069553088 @default.
• W3010420410 creator A5070191030 @default.
• W3010420410 creator A5076900528 @default.
• W3010420410 date "2019-06-27" @default.
• W3010420410 modified "2023-10-18" @default.
• W3010420410 title "Hard Pixel Mining for Depth Privileged Semantic Segmentation" @default.
• W3010420410 cites W125693051 @default.
• W3010420410 cites W1686810756 @default.
• W3010420410 cites W1821462560 @default.
• W3010420410 cites W1901129140 @default.
• W3010420410 cites W1903029394 @default.
• W3010420410 cites W1905829557 @default.
• W3010420410 cites W1915250530 @default.
• W3010420410 cites W1923115158 @default.
• W3010420410 cites W1923184257 @default.
• W3010420410 cites W1924762813 @default.
• W3010420410 cites W1938976761 @default.
• W3010420410 cites W1969366022 @default.
• W3010420410 cites W1972888803 @default.
• W3010420410 cites W2022508996 @default.
• W3010420410 cites W2067912884 @default.
• W3010420410 cites W2079188124 @default.
• W3010420410 cites W2100904656 @default.
• W3010420410 cites W2103328396 @default.
• W3010420410 cites W2111077768 @default.
• W3010420410 cites W2118181058 @default.
• W3010420410 cites W2124592697 @default.
• W3010420410 cites W2126942721 @default.
• W3010420410 cites W2132984949 @default.
• W3010420410 cites W2136331731 @default.
• W3010420410 cites W2161236525 @default.
• W3010420410 cites W2183341477 @default.
• W3010420410 cites W2194775991 @default.
• W3010420410 cites W2253986341 @default.
• W3010420410 cites W2264432461 @default.
• W3010420410 cites W2296073425 @default.
• W3010420410 cites W2325835689 @default.
• W3010420410 cites W2333621733 @default.
• W3010420410 cites W2339172515 @default.
• W3010420410 cites W2343077198 @default.
• W3010420410 cites W2363792106 @default.
• W3010420410 cites W2412782625 @default.
• W3010420410 cites W2463402750 @default.
• W3010420410 cites W2473131906 @default.
• W3010420410 cites W2485522583 @default.
• W3010420410 cites W2489780108 @default.
• W3010420410 cites W2560023338 @default.
• W3010420410 cites W2563705555 @default.
• W3010420410 cites W2587989515 @default.
• W3010420410 cites W2592165076 @default.
• W3010420410 cites W2606302360 @default.
• W3010420410 cites W2609822318 @default.
• W3010420410 cites W2630837129 @default.
• W3010420410 cites W2736374171 @default.
• W3010420410 cites W2739759330 @default.
• W3010420410 cites W2742400085 @default.
• W3010420410 cites W2768924212 @default.
• W3010420410 cites W2775906317 @default.
• W3010420410 cites W2776622059 @default.
• W3010420410 cites W2777686015 @default.
• W3010420410 cites W2793404847 @default.
• W3010420410 cites W2798791840 @default.
• W3010420410 cites W2809440904 @default.
• W3010420410 cites W2889002172 @default.
• W3010420410 cites W2889831618 @default.
• W3010420410 cites W2893890202 @default.
• W3010420410 cites W2895401575 @default.
• W3010420410 cites W2908138940 @default.
• W3010420410 cites W2925651774 @default.
• W3010420410 cites W2929754429 @default.
• W3010420410 cites W2946081696 @default.
• W3010420410 cites W2949086864 @default.
• W3010420410 cites W2951234442 @default.
• W3010420410 cites W2959581809 @default.
• W3010420410 cites W2962864714 @default.
• W3010420410 cites W2963052201 @default.
• W3010420410 cites W2963351448 @default.
• W3010420410 cites W2963368804 @default.
• W3010420410 cites W2963516811 @default.
• W3010420410 cites W2963896186 @default.
• W3010420410 cites W2964145160 @default.
• W3010420410 cites W2964269771 @default.
• W3010420410 cites W2964309882 @default.
• W3010420410 cites W2964327384 @default.
• W3010420410 cites W2968136324 @default.
• W3010420410 cites W385190221 @default.
• W3010420410 cites W3023820860 @default.
• W3010420410 doi "https://doi.org/10.48550/arxiv.1906.11437" @default.
• W3010420410 hasPublicationYear "2019" @default.
• W3010420410 type Work @default.
• W3010420410 sameAs 3010420410 @default.
• W3010420410 citedByCount "0" @default.
• W3010420410 crossrefType "posted-content" @default.
• W3010420410 hasAuthorship W3010420410A5032618817 @default.
• W3010420410 hasAuthorship W3010420410A5069553088 @default.
• W3010420410 hasAuthorship W3010420410A5070191030 @default.
• W3010420410 hasAuthorship W3010420410A5076900528 @default.

• next