Matches in SemOpenAlex for { <https://semopenalex.org/work/W4386065718> ?p ?o ?g. }
- W4386065718 abstract "We consider the generic problem of detecting low-level structures in images, which includes segmenting the manipulated parts, identifying out-of-focus pixels, separating shadow regions, and detecting concealed objects. Whereas each such topic has been typically addressed with a domain-specific solution, we show that a unified approach performs well across all of them. We take inspiration from the widely-used pre-training and then prompt tuning protocols in NLP and propose a new visual prompting model, named Explicit Visual Prompting (EVP). Different from the previous visual prompting which is typically a dataset-level implicit embedding, our key insight is to enforce the tunable parameters focusing on the explicit visual content from each individual image, i.e., the features from frozen patch embeddings and the input's high-frequency components. The proposed EVP significantly outperforms other parameter-efficient tuning protocols under the same amount of tunable parameters (5.7% extra trainable parameters of each task). EVP also achieves state-of-the-art performances on diverse low-level structure segmentation tasks compared to task-specific solutions. Our code is available at: https://github.com/NiFangBaAGe/Explicit-Visual-Prompt." @default.
- W4386065718 created "2023-08-23" @default.
- W4386065718 creator A5005772506 @default.
- W4386065718 creator A5050945995 @default.
- W4386065718 creator A5058799911 @default.
- W4386065718 creator A5082966443 @default.
- W4386065718 date "2023-06-01" @default.
- W4386065718 modified "2023-10-06" @default.
- W4386065718 title "Explicit Visual Prompting for Low-Level Structure Segmentations" @default.
- W4386065718 cites W1572666543 @default.
- W4386065718 cites W1884470205 @default.
- W4386065718 cites W1916731006 @default.
- W4386065718 cites W2009130368 @default.
- W4386065718 cites W2045584607 @default.
- W4386065718 cites W2059907896 @default.
- W4386065718 cites W2073839959 @default.
- W4386065718 cites W2085910047 @default.
- W4386065718 cites W2092792357 @default.
- W4386065718 cites W2097506049 @default.
- W4386065718 cites W2108598243 @default.
- W4386065718 cites W2121029817 @default.
- W4386065718 cites W2123302273 @default.
- W4386065718 cites W2130210787 @default.
- W4386065718 cites W2131297486 @default.
- W4386065718 cites W2141562793 @default.
- W4386065718 cites W2143373691 @default.
- W4386065718 cites W2153361333 @default.
- W4386065718 cites W2166502676 @default.
- W4386065718 cites W2169439181 @default.
- W4386065718 cites W2194775991 @default.
- W4386065718 cites W2292934097 @default.
- W4386065718 cites W2519623608 @default.
- W4386065718 cites W2566605673 @default.
- W4386065718 cites W2604933682 @default.
- W4386065718 cites W2752015292 @default.
- W4386065718 cites W2771617895 @default.
- W4386065718 cites W2799247888 @default.
- W4386065718 cites W2800623305 @default.
- W4386065718 cites W2884217841 @default.
- W4386065718 cites W2895126795 @default.
- W4386065718 cites W2943545929 @default.
- W4386065718 cites W2948407537 @default.
- W4386065718 cites W2948670693 @default.
- W4386065718 cites W2948988617 @default.
- W4386065718 cites W2954432404 @default.
- W4386065718 cites W2963014378 @default.
- W4386065718 cites W2963447079 @default.
- W4386065718 cites W2963819848 @default.
- W4386065718 cites W2964146055 @default.
- W4386065718 cites W2998154526 @default.
- W4386065718 cites W3027377338 @default.
- W4386065718 cites W3034627419 @default.
- W4386065718 cites W3034684132 @default.
- W4386065718 cites W3034779438 @default.
- W4386065718 cites W3108292056 @default.
- W4386065718 cites W3136675246 @default.
- W4386065718 cites W3138516171 @default.
- W4386065718 cites W3168112135 @default.
- W4386065718 cites W3170841864 @default.
- W4386065718 cites W3176152216 @default.
- W4386065718 cites W3179443972 @default.
- W4386065718 cites W3183637664 @default.
- W4386065718 cites W3191117306 @default.
- W4386065718 cites W4231972595 @default.
- W4386065718 cites W4243246777 @default.
- W4386065718 cites W4255023712 @default.
- W4386065718 cites W4312933868 @default.
- W4386065718 doi "https://doi.org/10.1109/cvpr52729.2023.01862" @default.
- W4386065718 hasPublicationYear "2023" @default.
- W4386065718 type Work @default.
- W4386065718 citedByCount "3" @default.
- W4386065718 countsByYear W43860657182023 @default.
- W4386065718 crossrefType "proceedings-article" @default.
- W4386065718 hasAuthorship W4386065718A5005772506 @default.
- W4386065718 hasAuthorship W4386065718A5050945995 @default.
- W4386065718 hasAuthorship W4386065718A5058799911 @default.
- W4386065718 hasAuthorship W4386065718A5082966443 @default.
- W4386065718 hasConcept C111919701 @default.
- W4386065718 hasConcept C120665830 @default.
- W4386065718 hasConcept C121332964 @default.
- W4386065718 hasConcept C134306372 @default.
- W4386065718 hasConcept C153180895 @default.
- W4386065718 hasConcept C154945302 @default.
- W4386065718 hasConcept C160633673 @default.
- W4386065718 hasConcept C162324750 @default.
- W4386065718 hasConcept C177264268 @default.
- W4386065718 hasConcept C187736073 @default.
- W4386065718 hasConcept C192209626 @default.
- W4386065718 hasConcept C199360897 @default.
- W4386065718 hasConcept C26517878 @default.
- W4386065718 hasConcept C2776760102 @default.
- W4386065718 hasConcept C2780451532 @default.
- W4386065718 hasConcept C31972630 @default.
- W4386065718 hasConcept C33923547 @default.
- W4386065718 hasConcept C36464697 @default.
- W4386065718 hasConcept C36503486 @default.
- W4386065718 hasConcept C38652104 @default.
- W4386065718 hasConcept C41008148 @default.
- W4386065718 hasConcept C41608201 @default.
- W4386065718 hasConcept C43126263 @default.