Matches in SemOpenAlex for { <https://semopenalex.org/work/W4225506590> ?p ?o ?g. }
Showing items 1 to 82 of
82
with 100 items per page.
- W4225506590 abstract "Recently, zero-shot image classification by vision-language pre-training has demonstrated incredible achievements, that the model can classify arbitrary category without seeing additional annotated images of that category. However, it is still unclear how to make the zero-shot recognition working well on broader vision problems, such as object detection and semantic segmentation. In this paper, we target for zero-shot semantic segmentation, by building it on an off-the-shelf pre-trained vision-language model, i.e., CLIP. It is difficult because semantic segmentation and the CLIP model perform on different visual granularity, that semantic segmentation processes on pixels while CLIP performs on images. To remedy the discrepancy on processing granularity, we refuse the use of the prevalent one-stage FCN based framework, and advocate a two-stage semantic segmentation framework, with the first stage extracting generalizable mask proposals and the second stage leveraging an image based CLIP model to perform zero-shot classification on the masked image crops which are generated in the first stage. Our experimental results show that this simple framework surpasses previous state-of-the-arts by a large margin: +29.5 hIoU on the Pascal VOC 2012 dataset, and +8.9 hIoU on the COCO Stuff dataset. With its simplicity and strong performance, we hope this framework to serve as a baseline to facilitate the future research." @default.
- W4225506590 created "2022-05-05" @default.
- W4225506590 creator A5004662803 @default.
- W4225506590 creator A5014838804 @default.
- W4225506590 creator A5039363991 @default.
- W4225506590 creator A5053949969 @default.
- W4225506590 creator A5062247384 @default.
- W4225506590 creator A5088888083 @default.
- W4225506590 creator A5090973869 @default.
- W4225506590 date "2021-12-29" @default.
- W4225506590 modified "2023-09-23" @default.
- W4225506590 title "A Simple Baseline for Zero-shot Semantic Segmentation with Pre-trained Vision-language Model" @default.
- W4225506590 hasPublicationYear "2021" @default.
- W4225506590 type Work @default.
- W4225506590 citedByCount "0" @default.
- W4225506590 crossrefType "posted-content" @default.
- W4225506590 hasAuthorship W4225506590A5004662803 @default.
- W4225506590 hasAuthorship W4225506590A5014838804 @default.
- W4225506590 hasAuthorship W4225506590A5039363991 @default.
- W4225506590 hasAuthorship W4225506590A5053949969 @default.
- W4225506590 hasAuthorship W4225506590A5062247384 @default.
- W4225506590 hasAuthorship W4225506590A5088888083 @default.
- W4225506590 hasAuthorship W4225506590A5090973869 @default.
- W4225506590 hasBestOaLocation W42255065901 @default.
- W4225506590 hasConcept C111368507 @default.
- W4225506590 hasConcept C111472728 @default.
- W4225506590 hasConcept C111919701 @default.
- W4225506590 hasConcept C119857082 @default.
- W4225506590 hasConcept C12725497 @default.
- W4225506590 hasConcept C127313418 @default.
- W4225506590 hasConcept C138885662 @default.
- W4225506590 hasConcept C153180895 @default.
- W4225506590 hasConcept C154945302 @default.
- W4225506590 hasConcept C177774035 @default.
- W4225506590 hasConcept C178790620 @default.
- W4225506590 hasConcept C185592680 @default.
- W4225506590 hasConcept C199360897 @default.
- W4225506590 hasConcept C204321447 @default.
- W4225506590 hasConcept C2776372474 @default.
- W4225506590 hasConcept C2778344882 @default.
- W4225506590 hasConcept C31972630 @default.
- W4225506590 hasConcept C41008148 @default.
- W4225506590 hasConcept C75608658 @default.
- W4225506590 hasConcept C774472 @default.
- W4225506590 hasConcept C89600930 @default.
- W4225506590 hasConceptScore W4225506590C111368507 @default.
- W4225506590 hasConceptScore W4225506590C111472728 @default.
- W4225506590 hasConceptScore W4225506590C111919701 @default.
- W4225506590 hasConceptScore W4225506590C119857082 @default.
- W4225506590 hasConceptScore W4225506590C12725497 @default.
- W4225506590 hasConceptScore W4225506590C127313418 @default.
- W4225506590 hasConceptScore W4225506590C138885662 @default.
- W4225506590 hasConceptScore W4225506590C153180895 @default.
- W4225506590 hasConceptScore W4225506590C154945302 @default.
- W4225506590 hasConceptScore W4225506590C177774035 @default.
- W4225506590 hasConceptScore W4225506590C178790620 @default.
- W4225506590 hasConceptScore W4225506590C185592680 @default.
- W4225506590 hasConceptScore W4225506590C199360897 @default.
- W4225506590 hasConceptScore W4225506590C204321447 @default.
- W4225506590 hasConceptScore W4225506590C2776372474 @default.
- W4225506590 hasConceptScore W4225506590C2778344882 @default.
- W4225506590 hasConceptScore W4225506590C31972630 @default.
- W4225506590 hasConceptScore W4225506590C41008148 @default.
- W4225506590 hasConceptScore W4225506590C75608658 @default.
- W4225506590 hasConceptScore W4225506590C774472 @default.
- W4225506590 hasConceptScore W4225506590C89600930 @default.
- W4225506590 hasLocation W42255065901 @default.
- W4225506590 hasOpenAccess W4225506590 @default.
- W4225506590 hasPrimaryLocation W42255065901 @default.
- W4225506590 hasRelatedWork W10029978 @default.
- W4225506590 hasRelatedWork W10828093 @default.
- W4225506590 hasRelatedWork W12546350 @default.
- W4225506590 hasRelatedWork W1469282 @default.
- W4225506590 hasRelatedWork W1865761 @default.
- W4225506590 hasRelatedWork W274842 @default.
- W4225506590 hasRelatedWork W292799 @default.
- W4225506590 hasRelatedWork W4560741 @default.
- W4225506590 hasRelatedWork W6930659 @default.
- W4225506590 hasRelatedWork W7003802 @default.
- W4225506590 isParatext "false" @default.
- W4225506590 isRetracted "false" @default.
- W4225506590 workType "article" @default.