Matches in SemOpenAlex for { <https://semopenalex.org/work/W4382246151> ?p ?o ?g. }
Showing items 1 to 52 of
52
with 100 items per page.
- W4382246151 abstract "We introduce episode-based training into zero-shot semantic segmentation (ZS3) for the first time. In particular, the model is trained on a set of simulated ZS3 tasks. The model gains the ability to predict simulated unseen classes over multiple episodes, which generalizes well to true unseen classes after training on multiple episodes. On the basis of this training framework, we propose a visual semantic alignment network named VSAN as our basic model, which mainly includes a feature extractor and semantic projection network. This base model constrains the visual-semantic distribution of the same class by a distance measure between visual features and semantic prototypes. In the inference phase, the trained projection network can generate corresponding semantic prototypes for all classes, and predict the segmentation results of the entire image by measuring the distance between visual features and semantic prototypes. The base model VSAN is called EB-VSAN after using the episode-based training strategy. Our model is a discriminative model, as opposed to generative methods, our model does not need retraining the classifier when a new class emerges, while avoiding multi-stage training. Our extensive ZS3 experiments on the benchmark dataset show that the EB-VSAN model outperforms current state-of-the-art methods, specifically, our hIoU metric outperforms state-of-the-art methods by an average of 1.7% on PASCAL VOC, while Average 2.5% improvement on PASCAL Context." @default.
- W4382246151 created "2023-06-28" @default.
- W4382246151 creator A5019393431 @default.
- W4382246151 creator A5032543777 @default.
- W4382246151 creator A5060006875 @default.
- W4382246151 date "2023-06-27" @default.
- W4382246151 modified "2023-10-16" @default.
- W4382246151 title "Episode-based training strategy for zero-shot semantic segmentation" @default.
- W4382246151 doi "https://doi.org/10.1117/12.2680262" @default.
- W4382246151 hasPublicationYear "2023" @default.
- W4382246151 type Work @default.
- W4382246151 citedByCount "0" @default.
- W4382246151 crossrefType "proceedings-article" @default.
- W4382246151 hasAuthorship W4382246151A5019393431 @default.
- W4382246151 hasAuthorship W4382246151A5032543777 @default.
- W4382246151 hasAuthorship W4382246151A5060006875 @default.
- W4382246151 hasConcept C119857082 @default.
- W4382246151 hasConcept C153180895 @default.
- W4382246151 hasConcept C154945302 @default.
- W4382246151 hasConcept C199360897 @default.
- W4382246151 hasConcept C2776214188 @default.
- W4382246151 hasConcept C41008148 @default.
- W4382246151 hasConcept C75608658 @default.
- W4382246151 hasConcept C89600930 @default.
- W4382246151 hasConcept C95623464 @default.
- W4382246151 hasConcept C97931131 @default.
- W4382246151 hasConceptScore W4382246151C119857082 @default.
- W4382246151 hasConceptScore W4382246151C153180895 @default.
- W4382246151 hasConceptScore W4382246151C154945302 @default.
- W4382246151 hasConceptScore W4382246151C199360897 @default.
- W4382246151 hasConceptScore W4382246151C2776214188 @default.
- W4382246151 hasConceptScore W4382246151C41008148 @default.
- W4382246151 hasConceptScore W4382246151C75608658 @default.
- W4382246151 hasConceptScore W4382246151C89600930 @default.
- W4382246151 hasConceptScore W4382246151C95623464 @default.
- W4382246151 hasConceptScore W4382246151C97931131 @default.
- W4382246151 hasLocation W43822461511 @default.
- W4382246151 hasOpenAccess W4382246151 @default.
- W4382246151 hasPrimaryLocation W43822461511 @default.
- W4382246151 hasRelatedWork W1972656095 @default.
- W4382246151 hasRelatedWork W2024160000 @default.
- W4382246151 hasRelatedWork W2061273563 @default.
- W4382246151 hasRelatedWork W2285052147 @default.
- W4382246151 hasRelatedWork W2729514902 @default.
- W4382246151 hasRelatedWork W2743258233 @default.
- W4382246151 hasRelatedWork W2773500201 @default.
- W4382246151 hasRelatedWork W2970216048 @default.
- W4382246151 hasRelatedWork W2998168123 @default.
- W4382246151 hasRelatedWork W4287995534 @default.
- W4382246151 isParatext "false" @default.
- W4382246151 isRetracted "false" @default.
- W4382246151 workType "article" @default.