- Browse by Subject
Browsing by Subject "Semantics"
Now showing 1 - 2 of 2
Results Per Page
Sort Options
Item Vagueness and Its Boundaries: A Peircean Theory of Vagueness(2010-02-26T18:35:42Z) Agler, David Wells; De Waal, Cornelis; De Tienne, André; Houser, NathanMany theories of vagueness employ question-begging assumptions about the semantic boundaries between truth and falsity. This thesis defends a theory of vagueness put forward by Charles S. Peirce and argues for a novel solution to the sorites paradox based upon his work. Contrary to widespread opinion, I argue that Peirce distinguished borderline vagueness from other related forms of indeterminacy, e.g. indefiniteness, generality, unspecificity, uninformativity, etc. By clarifying Peirce’s conception of borderline vagueness, I argue for a solution to the sorites paradox based upon his logical semantics. In addition, I argue for this theory against the epistemic theory of vagueness, which makes controversial claims concerning the sharp semantic boundary between truth and falsity, and against the supervaluationist theory of vagueness, which is committed to the in principle impossibility of sharp semantics boundaries for propositions with vague terms.Item Weakly-Supervised Cross-Domain Adaptation for Endoscopic Lesions Segmentation(IEEE, 2021) Dong, Jiahua; Cong, Yang; Sun, Gan; Yang, Yunsheng; Xu, Xiaowei; Ding, Zhengming; Computer Information and Graphics Technology, School of Engineering and TechnologyWeakly-supervised learning has attracted growing research attention on medical lesions segmentation due to significant saving in pixel-level annotation cost. However, 1) most existing methods require effective prior and constraints to explore the intrinsic lesions characterization, which only generates incorrect and rough prediction; 2) they neglect the underlying semantic dependencies among weakly-labeled target enteroscopy diseases and fully-annotated source gastroscope lesions, while forcefully utilizing untransferable dependencies leads to the negative performance. To tackle above issues, we propose a new weakly-supervised lesions transfer framework, which can not only explore transferable domain-invariant knowledge across different datasets, but also prevent the negative transfer of untransferable representations. Specifically, a Wasserstein quantified transferability framework is developed to highlight wide-range transferable contextual dependencies, while neglecting the irrelevant semantic characterizations. Moreover, a novel self-supervised pseudo label generator is designed to equally provide confident pseudo pixel labels for both hard-to-transfer and easy-to-transfer target samples. It inhibits the enormous deviation of false pseudo pixel labels under the self-supervision manner. Afterwards, dynamically-searched feature centroids are aligned to narrow category-wise distribution shift. Comprehensive theoretical analysis and experiments show the superiority of our model on the endoscopic dataset and several public datasets.