Matches in SemOpenAlex for { <https://semopenalex.org/work/W3110109236> ?p ?o ?g. }
- W3110109236 endingPage "18" @default.
- W3110109236 startingPage "1" @default.
- W3110109236 abstract "Single-stage instance segmentation approaches have recently gained popularity due to their speed and simplicity, but are still lagging behind in accuracy, compared to two-stage methods. We propose a fast single-stage instance segmentation method, called SipMask, that preserves instance-specific spatial information by separating mask prediction of an instance to different sub-regions of a detected bounding-box. Our main contribution is a novel light-weight spatial preservation (SP) module that generates a separate set of spatial coefficients for each sub-region within a bounding-box, leading to improved mask predictions. It also enables accurate delineation of spatially adjacent instances. Further, we introduce a mask alignment weighting loss and a feature alignment scheme to better correlate mask prediction with object detection. On COCO test-dev, our SipMask outperforms the existing single-stage methods. Compared to the state-of-the-art single-stage TensorMask, SipMask obtains an absolute gain of 1.0% (mask AP), while providing a four-fold speedup. In terms of real-time capabilities, SipMask outperforms YOLACT with an absolute gain of 3.0% (mask AP) under similar settings, while operating at comparable speed on a Titan Xp. We also evaluate our SipMask for real-time video instance segmentation, achieving promising results on YouTube-VIS dataset. The source code is available at https://github.com/JialeCao001/SipMask ." @default.
- W3110109236 created "2020-12-07" @default.
- W3110109236 creator A5009362997 @default.
- W3110109236 creator A5022929869 @default.
- W3110109236 creator A5037129512 @default.
- W3110109236 creator A5079112987 @default.
- W3110109236 creator A5082634513 @default.
- W3110109236 creator A5086887025 @default.
- W3110109236 date "2020-01-01" @default.
- W3110109236 modified "2023-09-28" @default.
- W3110109236 title "SipMask: Spatial Information Preservation for Fast Image and Video Instance Segmentation" @default.
- W3110109236 cites W1765260417 @default.
- W3110109236 cites W1861492603 @default.
- W3110109236 cites W1903029394 @default.
- W3110109236 cites W2102605133 @default.
- W3110109236 cites W2117539524 @default.
- W3110109236 cites W2194775991 @default.
- W3110109236 cites W2216125271 @default.
- W3110109236 cites W2317851288 @default.
- W3110109236 cites W2322480645 @default.
- W3110109236 cites W2412782625 @default.
- W3110109236 cites W2555182955 @default.
- W3110109236 cites W2558156561 @default.
- W3110109236 cites W2565639579 @default.
- W3110109236 cites W2601564443 @default.
- W3110109236 cites W2608858501 @default.
- W3110109236 cites W2777795072 @default.
- W3110109236 cites W2894299524 @default.
- W3110109236 cites W2916797271 @default.
- W3110109236 cites W2920326761 @default.
- W3110109236 cites W2953433552 @default.
- W3110109236 cites W2962825871 @default.
- W3110109236 cites W2963150697 @default.
- W3110109236 cites W2963775509 @default.
- W3110109236 cites W2963849369 @default.
- W3110109236 cites W2963857746 @default.
- W3110109236 cites W2963927307 @default.
- W3110109236 cites W2964236837 @default.
- W3110109236 cites W2966926453 @default.
- W3110109236 cites W2979579363 @default.
- W3110109236 cites W2981537222 @default.
- W3110109236 cites W2981927700 @default.
- W3110109236 cites W2982103617 @default.
- W3110109236 cites W2982161360 @default.
- W3110109236 cites W2982723417 @default.
- W3110109236 cites W2982770724 @default.
- W3110109236 cites W2985276900 @default.
- W3110109236 cites W2985405845 @default.
- W3110109236 cites W2986357608 @default.
- W3110109236 cites W2990075400 @default.
- W3110109236 cites W2990361805 @default.
- W3110109236 cites W2993182889 @default.
- W3110109236 cites W2997999719 @default.
- W3110109236 cites W2998792609 @default.
- W3110109236 cites W3034522582 @default.
- W3110109236 cites W3034681942 @default.
- W3110109236 cites W3034826836 @default.
- W3110109236 cites W3034934229 @default.
- W3110109236 cites W3034955056 @default.
- W3110109236 cites W3035709993 @default.
- W3110109236 cites W3106732900 @default.
- W3110109236 cites W3107653507 @default.
- W3110109236 doi "https://doi.org/10.1007/978-3-030-58568-6_1" @default.
- W3110109236 hasPublicationYear "2020" @default.
- W3110109236 type Work @default.
- W3110109236 sameAs 3110109236 @default.
- W3110109236 citedByCount "51" @default.
- W3110109236 countsByYear W31101092362021 @default.
- W3110109236 countsByYear W31101092362022 @default.
- W3110109236 countsByYear W31101092362023 @default.
- W3110109236 crossrefType "book-chapter" @default.
- W3110109236 hasAuthorship W3110109236A5009362997 @default.
- W3110109236 hasAuthorship W3110109236A5022929869 @default.
- W3110109236 hasAuthorship W3110109236A5037129512 @default.
- W3110109236 hasAuthorship W3110109236A5079112987 @default.
- W3110109236 hasAuthorship W3110109236A5082634513 @default.
- W3110109236 hasAuthorship W3110109236A5086887025 @default.
- W3110109236 hasBestOaLocation W31101092362 @default.
- W3110109236 hasConcept C115961682 @default.
- W3110109236 hasConcept C124504099 @default.
- W3110109236 hasConcept C126838900 @default.
- W3110109236 hasConcept C138885662 @default.
- W3110109236 hasConcept C147037132 @default.
- W3110109236 hasConcept C153180895 @default.
- W3110109236 hasConcept C154945302 @default.
- W3110109236 hasConcept C173608175 @default.
- W3110109236 hasConcept C183115368 @default.
- W3110109236 hasConcept C2776151529 @default.
- W3110109236 hasConcept C2776401178 @default.
- W3110109236 hasConcept C31972630 @default.
- W3110109236 hasConcept C41008148 @default.
- W3110109236 hasConcept C41895202 @default.
- W3110109236 hasConcept C63584917 @default.
- W3110109236 hasConcept C68339613 @default.
- W3110109236 hasConcept C71924100 @default.
- W3110109236 hasConcept C89600930 @default.
- W3110109236 hasConceptScore W3110109236C115961682 @default.
- W3110109236 hasConceptScore W3110109236C124504099 @default.