DC Field | Value | Language |
---|---|---|
dc.contributor.author | Song, Hwanjun | ko |
dc.contributor.author | Bang, Jihwan | ko |
dc.date.accessioned | 2024-08-29T02:00:07Z | - |
dc.date.available | 2024-08-29T02:00:07Z | - |
dc.date.created | 2024-06-11 | - |
dc.date.issued | 2024-11 | - |
dc.identifier.citation | PATTERN RECOGNITION, v.155 | - |
dc.identifier.issn | 0031-3203 | - |
dc.identifier.uri | http://hdl.handle.net/10203/322447 | - |
dc.description.abstract | Prompt-OVD is an efficient and effective DETR-based framework for open -vocabulary object detection that utilizes class embeddings from CLIP as prompts, guiding the Transformer decoder to detect objects in base and novel classes. Additionally, our RoI-pruned masked attention helps leverage the zero -shot classification ability of the Vision Transformer -based CLIP, resulting in improved detection performance at a minimal computational cost. Our experiments on the OV-COCO and OV-LVIS datasets demonstrate that Prompt-OVD achieves an impressive 21.2 times faster inference speed than the first end -to -end open -vocabulary detection method (OVDETR), while also achieving higher APs than four two -stage methods operating within similar inference time ranges. We release the code at https://github.com/DISL-Lab/Prompt-OVD. | - |
dc.language | English | - |
dc.publisher | ELSEVIER SCI LTD | - |
dc.title | Prompt-guided DETR with RoI-pruned masked attention for open-vocabulary object detection | - |
dc.type | Article | - |
dc.identifier.wosid | 001257471900001 | - |
dc.identifier.scopusid | 2-s2.0-85195806366 | - |
dc.type.rims | ART | - |
dc.citation.volume | 155 | - |
dc.citation.publicationname | PATTERN RECOGNITION | - |
dc.identifier.doi | 10.1016/j.patcog.2024.110648 | - |
dc.contributor.localauthor | Song, Hwanjun | - |
dc.description.isOpenAccess | N | - |
dc.type.journalArticle | Article | - |
dc.subject.keywordAuthor | OVD | - |
dc.subject.keywordAuthor | Transformer | - |
dc.subject.keywordAuthor | Object detection | - |
dc.subject.keywordAuthor | Open-vocabulary detection | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.