Global ETD Search
Search theses and dissertations gathered from participating repositories worldwide. Every result links back to the library that holds it. No account is needed.
Results
Showing 1 to 9 of 9 for “"consistency training"”.
-
Semi-supervised cycle-consistency training for end-to-end ASR using unpaired speech
… of limited paired data, the idea of cycle-consistency losses has been proposed recently in areas such as machine translation and computer vision. In ASR, cycle-consistency training is achieved by building a reverse system, e.g., a text-to-speech system, and designing a loss based on the …
-
From Phishing URL Generation to Cross-Domain Fake News Detection: Data-Efficient Methods for Online Security
… attacks spread quickly, but labels used for training detection systems typically lag behind new events. As a result, text classifiers deployed in real-world settings frequently face performance drops due to changes in topics, language usage, and attacker methods, known as distribution shift. …
-
Class consistency learning : training instance segmentation models to recognize object co-occurrence and class alignment
… segmentation models can learn \textit{class consistency}, that is, to identify common contextual or relational patterns of objects within a given image and harness that information to improve detection performance in industrial applications. To enable this, we developed a custom grid-format …
-
Strategies for Modulating Spinal Movement Patterns and Variability
… visual cues into an exploratory movement training program aiming to increase movement variability and a movement consistency training program aiming to decrease movement variability in a healthy population. The results demonstrated that one session of a training program aiming to modulate …
-
Learning Object Detection with Weak Supervision
… in many computer vision applications. However, training deep models typically requires large-scale datasets with elaborate annotations. Collecting and annotating large-scale datasets are laborious, especially for object detection --- a challenging vision task. A promising solution for reducing …
-
Robust White Matter Hyperintensity Segmentation
… a different center than the one used for model training can result in an inaccurate segmentation.<br/><br/>In this thesis, I proposed different approaches to tackle the aforementioned challenges. Firstly, I address inter modality-heterogeneity; I proposed an efficient strategy to leverage …
-
Advanced Radar Sounder Data Analysis Methods under Limited labeled Data Constraints
… collections while leveraging semi-supervised consistency training. This approach significantly reduces annotation costs and accelerates large-scale subsurface mapping. The third contribution focuses on introducing the use of foundation models for the analysis of radar sounder data to improve …
-
Information Extraction for Low-Resource Schemas
… and calibration. We first discuss structural consistency, a new measure of robustness in information extraction based on compositionality. We present structural consistency post-training (SCPT) as a data augmentation method to boost structural consistency for a wide range of model …